| blob | 3bd3a6f01cb8be567b3783ed9cc139f7bfb45b4d |
1 ;;; org-element.el --- Parser for Org Syntax -*- lexical-binding: t; -*-
3 ;; Copyright (C) 2012-2018 Free Software Foundation, Inc.
5 ;; Author: Nicolas Goaziou <n.goaziou at gmail dot com>
6 ;; Keywords: outlines, hypermedia, calendar, wp
8 ;; This file is part of GNU Emacs.
10 ;; GNU Emacs is free software: you can redistribute it and/or modify
11 ;; it under the terms of the GNU General Public License as published by
12 ;; the Free Software Foundation, either version 3 of the License, or
13 ;; (at your option) any later version.
15 ;; GNU Emacs is distributed in the hope that it will be useful,
16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 ;; GNU General Public License for more details.
20 ;; You should have received a copy of the GNU General Public License
21 ;; along with GNU Emacs. If not, see <https://www.gnu.org/licenses/>.
23 ;;; Commentary:
24 ;;
25 ;; See <https://orgmode.org/worg/dev/org-syntax.html> for details about
26 ;; Org syntax.
27 ;;
28 ;; Lisp-wise, a syntax object can be represented as a list.
29 ;; It follows the pattern (TYPE PROPERTIES CONTENTS), where:
30 ;; TYPE is a symbol describing the object.
31 ;; PROPERTIES is the property list attached to it. See docstring of
32 ;; appropriate parsing function to get an exhaustive list.
33 ;; CONTENTS is a list of syntax objects or raw strings contained
34 ;; in the current object, when applicable.
35 ;;
36 ;; For the whole document, TYPE is `org-data' and PROPERTIES is nil.
37 ;;
38 ;; The first part of this file defines constants for the Org syntax,
39 ;; while the second one provide accessors and setters functions.
40 ;;
41 ;; The next part implements a parser and an interpreter for each
42 ;; element and object type in Org syntax.
43 ;;
44 ;; The following part creates a fully recursive buffer parser. It
45 ;; also provides a tool to map a function to elements or objects
46 ;; matching some criteria in the parse tree. Functions of interest
47 ;; are `org-element-parse-buffer', `org-element-map' and, to a lesser
48 ;; extent, `org-element-parse-secondary-string'.
49 ;;
50 ;; The penultimate part is the cradle of an interpreter for the
51 ;; obtained parse tree: `org-element-interpret-data'.
52 ;;
53 ;; The library ends by furnishing `org-element-at-point' function, and
54 ;; a way to give information about document structure around point
55 ;; with `org-element-context'. A cache mechanism is also provided for
56 ;; these functions.
59 ;;; Code:
66 \f
67 ;;; Definitions And Rules
68 ;;
69 ;; Define elements, greater elements and specify recursive objects,
70 ;; along with the affiliated keywords recognized. Also set up
71 ;; restrictions on recursive objects combinations.
72 ;;
73 ;; `org-element-update-syntax' builds proper syntax regexps according
74 ;; to current setup.
77 "Regexp to separate paragraphs in an Org buffer.
79 is not sufficient to know if point is at a paragraph ending. See
83 "Regexp possibly matching the beginning of an object.
84 This regexp allows false positives. Dedicated parser (e.g.,
85 `org-export-bold-parser') will take care of further filtering.
86 Radio links are not matched by this regexp, as they are treated
90 "Build variable syntax regexps."
93 ;; Headlines, inlinetasks.
95 ;; Footnote definitions.
97 ;; Diary sexps.
100 ;; Empty lines.
102 ;; Tables (any type).
105 ;; Comments, keyword-like or block-like constructs.
106 ;; Blocks and keywords with dual values need to be
107 ;; double-checked.
112 ;; Drawers (any type) and fixed-width areas. Drawers
113 ;; need to be double-checked.
115 ;; Horizontal rules.
117 ;; LaTeX environments.
119 ;; Clock lines.
121 ;; Lists.
128 org-element--object-regexp
132 ;; Sub/superscript.
134 ;; Bold, code, italic, strike-through, underline
135 ;; and verbatim.
139 ;; Plain links.
141 ;; Objects starting with "[": regular link,
142 ;; footnote reference, statistics cookie,
143 ;; timestamp (inactive).
150 ;; Objects starting with "@": export snippets.
151 "@@"
152 ;; Objects starting with "{": macro.
153 "{{{"
154 ;; Objects starting with "<" : timestamp
155 ;; (active, diary), target, radio target and
156 ;; angular links.
158 ;; Objects starting with "$": latex fragment.
160 ;; Objects starting with "\": line break,
161 ;; entity, latex fragment.
163 ;; Objects starting with raw text: inline Babel
164 ;; source block, inline Babel call.
170 ;;;###autoload
172 "Update parser internals."
179 dynamic-block example-block export-block fixed-width
180 footnote-definition headline horizontal-rule inlinetask item
181 keyword latex-environment node-property paragraph plain-list
182 planning property-drawer quote-block section
183 special-block src-block table table-row verse-block)
188 item plain-list property-drawer quote-block section
189 special-block table)
194 inline-src-block italic line-break latex-fragment link macro
195 radio-target statistics-cookie strike-through subscript superscript
196 table-cell target timestamp underline verbatim)
201 superscript table-cell underline)
211 "List of affiliated keywords as strings.
219 "Alist of usual translations for keywords.
220 The key is the old name and the value the new one. The property
224 "List of affiliated keywords that can occur more than once in an element.
226 Their value will be consed into a list of strings, which will be
227 returned as the value of the property.
229 This list is checked after translations have been applied. See
230 `org-element-keyword-translation-alist'.
236 "List of affiliated keywords whose value can be parsed.
238 Their value will be stored as a secondary string: a list of
239 strings and objects.
241 This list is checked after translations have been applied. See
246 org-element-parsed-keywords)
247 "Alist of parsed keywords and associated properties.
248 This is generated from `org-element-parsed-keywords', which
252 "List of affiliated keywords which can have a secondary value.
254 In Org syntax, they can be written with optional square brackets
255 before the colons. For example, RESULTS keyword can be
256 associated to a hash value with the following:
258 #+RESULTS[hash-string]: some-source
260 This list is checked after translations have been applied. See
266 ;; Dual affiliated keywords.
270 ;; Regular affiliated keywords.
275 org-element-affiliated-keywords)))
277 ;; Export attributes.
279 "Regexp matching any affiliated keyword.
281 Keyword name is put in match group 1. Moreover, if keyword
282 belongs to `org-element-dual-keywords', put the dual value in
283 match group 2.
297 ;; Ignore all links in a link description. Also ignore
298 ;; radio-targets and line breaks.
300 italic latex-fragment macro statistics-cookie strike-through
301 subscript superscript underline verbatim)
303 ;; Remove any variable object from radio target as it would
304 ;; prevent it from being properly recognized.
306 subscript superscript underline superscript)
310 ;; Ignore inline babel call and inline src block as formulas are
311 ;; possible. Also ignore line breaks and statistics cookies.
313 latex-fragment link macro radio-target strike-through
314 subscript superscript target timestamp underline verbatim)
318 "Alist of objects restrictions.
320 key is an element or object type containing objects and value is
321 a list of types that can be contained within an element or object
322 of such type.
324 For example, in a `radio-target' object, one can only find
325 entities, latex-fragments, subscript, superscript and text
326 markup.
328 This alist also applies to secondary string. For example, an
329 `headline' type element doesn't directly contain objects, but
344 "Table used internally to pair only round brackets.
353 "Table used internally to pair only square brackets.
362 "Table used internally to pair only curly brackets.
366 "Parse paired brackets at point.
367 CHAR is the opening bracket to consider, as a character. Return
368 contents between brackets, as a string, or nil. Also move point
369 past the brackets."
384 \f
385 ;;; Accessors and Setters
386 ;;
387 ;; Provide four accessors: `org-element-type', `org-element-property'
388 ;; `org-element-contents' and `org-element-restriction'.
389 ;;
390 ;; Setter functions allow modification of elements by side effect.
391 ;; There is `org-element-put-property', `org-element-set-contents'.
392 ;; These low-level functions are useful to build a parse tree.
393 ;;
394 ;; `org-element-adopt-elements', `org-element-set-element',
395 ;; `org-element-extract-element' and `org-element-insert-before' are
396 ;; high-level functions useful to modify a parse tree.
397 ;;
398 ;; `org-element-secondary-p' is a predicate used to know if a given
399 ;; object belongs to a secondary string. `org-element-class' tells if
400 ;; some parsed data is an element or an object, handling pseudo
401 ;; elements and objects. `org-element-copy' returns an element or
402 ;; object, stripping its parent property in the process.
405 "Return type of ELEMENT.
407 The function returns the type of the element or object provided.
408 It can also return the following special value:
409 `plain-text' for a string
410 `org-data' for a complete document
411 nil in any other case."
417 "Extract the value from the PROPERTY of an ELEMENT."
422 "Extract contents from an ELEMENT."
428 "Return restriction associated to ELEMENT.
429 ELEMENT can be an element, an object or a symbol representing an
430 element or object type."
432 org-element-object-restrictions)))
435 "In ELEMENT set PROPERTY to VALUE.
436 Return modified element."
439 element))
442 "Set ELEMENT's contents to CONTENTS.
443 Return ELEMENT."
450 "Non-nil when OBJECT directly belongs to a secondary string.
451 Return value is the property name, as a keyword, or nil."
454 org-element-secondary-value-alist))))
461 "Return class for ELEMENT, as a symbol.
462 Class is either `element' or `object'. Optional argument PARENT
463 is the element or object containing DATUM. It defaults to the
464 value of DATUM `:parent' property."
468 ;; Trivial cases.
471 ;; Special cases.
475 ;; Pseudo object or elements. Make a guess about its class.
476 ;; Basically a pseudo object is contained within another object,
477 ;; a secondary string or a container element.
487 "Append elements to the contents of another element.
489 PARENT is an element or object. CHILDREN can be elements,
490 objects, or a strings.
492 The function takes care of setting `:parent' property for CHILD.
493 Return parent element."
496 ;; Link every child to PARENT. If PARENT is nil, it is a secondary
497 ;; string: parent is the list itself.
500 ;; Add CHILDREN at the end of PARENT contents.
503 parent
505 ;; Return modified PARENT element.
509 "Extract ELEMENT from parse tree.
510 Remove element from the parse tree by side-effect, and return it
511 with its `:parent' property stripped out."
516 parent secondary
519 parent
521 ;; Return ELEMENT with its :parent removed.
525 "Insert ELEMENT before LOCATION in parse tree.
526 LOCATION is an element, object or string within the parse tree.
527 Parse tree is modified by side effect."
532 ;; Special case: LOCATION is the first element of an
533 ;; independent secondary string (e.g. :title property). Add
534 ;; ELEMENT in-place.
538 ;; Install ELEMENT at the appropriate LOCATION within SIBLINGS.
547 ;; Store SIBLINGS at appropriate place in parse tree.
552 ;; Set appropriate :parent property.
556 "Replace element or object OLD with element or object NEW.
557 The function takes care of setting `:parent' property for NEW."
558 ;; Ensure OLD and NEW have the same parent.
562 ;; We cannot replace OLD with NEW since one of them is not an
563 ;; object or element. We take the long path.
566 ;; Since OLD is going to be changed into NEW by side-effect, first
567 ;; make sure that every element or object within NEW has OLD as
568 ;; parent.
571 ;; Transfer contents.
573 ;; Overwrite OLD's properties with NEW's.
575 ;; Transfer type.
579 "Create a new element of type TYPE.
580 Optional argument PROPS, when non-nil, is a plist defining the
581 properties of the element. CHILDREN can be elements, objects or
582 strings."
586 "Return a copy of DATUM.
587 DATUM is an element, object, string or nil. `:parent' property
588 is cleared and contents are removed in the process."
599 \f
600 ;;; Greater elements
601 ;;
602 ;; For each greater element type, we define a parser and an
603 ;; interpreter.
604 ;;
605 ;; A parser returns the element or object as the list described above.
606 ;; Most of them accepts no argument. Though, exceptions exist. Hence
607 ;; every element containing a secondary string (see
608 ;; `org-element-secondary-value-alist') will accept an optional
609 ;; argument to toggle parsing of these secondary strings. Moreover,
610 ;; `item' parser requires current list's structure as its first
611 ;; element.
612 ;;
613 ;; An interpreter accepts two arguments: the list representation of
614 ;; the element or object, and its contents. The latter may be nil,
615 ;; depending on the element or object considered. It returns the
616 ;; appropriate Org syntax, as a string.
617 ;;
618 ;; Parsing functions must follow the naming convention:
619 ;; org-element-TYPE-parser, where TYPE is greater element's type, as
620 ;; defined in `org-element-greater-elements'.
621 ;;
622 ;; Similarly, interpreting functions must follow the naming
623 ;; convention: org-element-TYPE-interpreter.
624 ;;
625 ;; With the exception of `headline' and `item' types, greater elements
626 ;; cannot contain other greater elements of their own type.
627 ;;
628 ;; Beside implementing a parser and an interpreter, adding a new
629 ;; greater element requires tweaking `org-element--current-element'.
630 ;; Moreover, the newly defined type must be added to both
631 ;; `org-element-all-elements' and `org-element-greater-elements'.
634 ;;;; Center Block
637 "Parse a center block.
639 LIMIT bounds the search. AFFILIATED is a list of which CAR is
640 the buffer position at the beginning of the first affiliated
641 keyword and CDR is a plist of affiliated keywords along with
642 their value.
644 Return a list whose CAR is `center-block' and CDR is a plist
645 containing `:begin', `:end', `:contents-begin', `:contents-end',
646 `:post-blank' and `:post-affiliated' keywords.
648 Assume point is at the beginning of the block."
652 ;; Incomplete block: parse it as a paragraph.
657 ;; Empty blocks have no contents.
671 :end end
672 :contents-begin contents-begin
673 :contents-end contents-end
679 "Interpret a center-block element as Org syntax.
680 CONTENTS is the contents of the element."
684 ;;;; Drawer
687 "Parse a drawer.
689 LIMIT bounds the search. AFFILIATED is a list of which CAR is
690 the buffer position at the beginning of the first affiliated
691 keyword and CDR is a plist of affiliated keywords along with
692 their value.
694 Return a list whose CAR is `drawer' and CDR is a plist containing
695 `:drawer-name', `:begin', `:end', `:contents-begin',
696 `:contents-end', `:post-blank' and `:post-affiliated' keywords.
698 Assume point is at beginning of drawer."
701 ;; Incomplete drawer: parse it as a paragraph.
709 ;; Empty drawers have no contents.
722 :end end
723 :drawer-name name
724 :contents-begin contents-begin
725 :contents-end contents-end
731 "Interpret DRAWER element as Org syntax.
732 CONTENTS is the contents of the element."
735 contents))
738 ;;;; Dynamic Block
741 "Parse a dynamic block.
743 LIMIT bounds the search. AFFILIATED is a list of which CAR is
744 the buffer position at the beginning of the first affiliated
745 keyword and CDR is a plist of affiliated keywords along with
746 their value.
748 Return a list whose CAR is `dynamic-block' and CDR is a plist
749 containing `:block-name', `:begin', `:end', `:contents-begin',
750 `:contents-end', `:arguments', `:post-blank' and
751 `:post-affiliated' keywords.
753 Assume point is at beginning of dynamic block."
757 ;; Incomplete block: parse it as a paragraph.
766 ;; Empty blocks have no contents.
779 :end end
780 :block-name name
781 :arguments arguments
782 :contents-begin contents-begin
783 :contents-end contents-end
789 "Interpret DYNAMIC-BLOCK element as Org syntax.
790 CONTENTS is the contents of the element."
795 contents))
798 ;;;; Footnote Definition
807 "Parse a footnote definition.
809 LIMIT bounds the search. AFFILIATED is a list of which CAR is
810 the buffer position at the beginning of the first affiliated
811 keyword and CDR is a plist of affiliated keywords along with
812 their value.
814 Return a list whose CAR is `footnote-definition' and CDR is
815 a plist containing `:label', `:begin' `:end', `:contents-begin',
816 `:contents-end', `:pre-blank',`:post-blank' and
817 `:post-affiliated' keywords.
819 Assume point is at the beginning of the footnote definition."
831 limit)
833 ;; At a new footnote definition, make sure we end
834 ;; before any affiliated keyword above.
860 :begin begin
861 :end end
862 :contents-begin contents-begin
864 :pre-blank pre-blank
870 "Interpret FOOTNOTE-DEFINITION element as Org syntax.
871 CONTENTS is the contents of the footnote-definition."
874 ;; 0 is specific to paragraphs at the beginning of
875 ;; the footnote definition, so we use 1 as
876 ;; a fall-back value, which is more universal.
878 ;; Footnote ends after more than two consecutive empty
879 ;; lines: limit ourselves to 2 newline characters.
886 ;;;; Headline
889 "Return node properties associated to headline at point.
890 Upcase property names. It avoids confusion between properties
891 obtained through property drawer and default properties from the
892 parser (e.g. `:end' and :END:). Return value is a plist."
904 properties))))
907 "Return time properties associated to headline at point.
908 Return value is a plist."
922 plist))))
925 "Parse a headline.
927 Return a list whose CAR is `headline' and CDR is a plist
928 containing `:raw-value', `:title', `:begin', `:end',
929 `:pre-blank', `:contents-begin' and `:contents-end', `:level',
930 `:priority', `:tags', `:todo-keyword',`:todo-type', `:scheduled',
931 `:deadline', `:closed', `:archivedp', `:commentedp'
932 `:footnote-section-p', `:post-blank' and `:post-affiliated'
933 keywords.
935 The plist also contains any property set in the property drawer,
936 with its name in upper cases and colons added at the
937 beginning (e.g., `:CUSTOM_ID').
939 LIMIT is a buffer position bounding the search.
941 When RAW-SECONDARY-P is non-nil, headline's title will not be
942 parsed as a secondary string, but as a plain string instead.
944 Assume point is at beginning of the headline."
990 :begin begin
991 :end end
992 :pre-blank
995 :contents-begin contents-begin
996 :contents-end contents-end
997 :level level
998 :priority priority
999 :tags tags
1000 :todo-keyword todo
1001 :todo-type todo-type
1002 :post-blank
1006 :footnote-section-p footnote-section-p
1007 :archivedp archivedp
1008 :commentedp commentedp
1010 time-props
1011 standard-props))))
1013 headline :title
1022 nil
1024 headline)))))))
1027 "Interpret HEADLINE element as Org syntax.
1028 CONTENTS is the contents of the element."
1041 ?*)
1045 " "
1048 org-footnote-section
1049 title))))
1051 heading
1052 ;; Align tags.
1060 ?\s)
1061 tags))
1065 tags))))
1067 contents)))
1070 ;;;; Inlinetask
1073 "Parse an inline task.
1075 Return a list whose CAR is `inlinetask' and CDR is a plist
1076 containing `:title', `:begin', `:end', `:pre-blank',
1077 `:contents-begin' and `:contents-end', `:level', `:priority',
1078 `:raw-value', `:tags', `:todo-keyword', `:todo-type',
1079 `:scheduled', `:deadline', `:closed', `:post-blank' and
1080 `:post-affiliated' keywords.
1082 The plist also contains any property set in the property drawer,
1083 with its name in upper cases and colons added at the
1084 beginning (e.g., `:CUSTOM_ID').
1086 When optional argument RAW-SECONDARY-P is non-nil, inline-task's
1087 title will not be parsed as a secondary string, but as a plain
1088 string instead.
1090 Assume point is at beginning of the inline task."
1137 :begin begin
1138 :end end
1139 :pre-blank
1142 :contents-begin contents-begin
1143 :contents-end contents-end
1144 :level level
1145 :priority priority
1146 :tags tags
1147 :todo-keyword todo
1148 :todo-type todo-type
1151 time-props
1152 standard-props))))
1154 inlinetask :title
1163 nil
1165 inlinetask))))))
1168 "Interpret INLINETASK element as Org syntax.
1169 CONTENTS is the contents of inlinetask."
1183 ;; Align tags.
1191 ?\s)
1192 tags))
1196 tags))))
1197 ;; Prefer degenerate inlinetasks when there are no
1198 ;; contents.
1201 contents
1205 ;;;; Item
1208 "Parse an item.
1210 STRUCT is the structure of the plain list.
1212 Return a list whose CAR is `item' and CDR is a plist containing
1213 `:bullet', `:begin', `:end', `:contents-begin', `:contents-end',
1214 `:checkbox', `:counter', `:tag', `:structure', `:pre-blank',
1215 `:post-blank' and `:post-affiliated' keywords.
1217 When optional argument RAW-SECONDARY-P is non-nil, item's tag, if
1218 any, will not be parsed as a secondary string, but as a plain
1219 string instead.
1221 Assume point is at the beginning of the item."
1246 ;; Ignore tags in un-ordered lists: they are just
1247 ;; a part of item's body.
1254 ;; If first line isn't empty, contents really
1255 ;; start at the text after item's meta-data.
1268 :begin begin
1269 :end end
1270 :contents-begin contents-begin
1271 :contents-end contents-end
1272 :checkbox checkbox
1273 :counter counter
1274 :structure struct
1275 :pre-blank pre-blank
1279 item :tag
1286 item))))))))
1289 "Interpret ITEM element as Org syntax.
1290 CONTENTS is the contents of the element."
1302 bullet
1311 tag
1316 ;; 0 is specific to paragraphs at the
1317 ;; beginning of the item, so we use 1 as
1318 ;; a fall-back value, which is more universal.
1320 ;; Lists ends after more than two consecutive
1321 ;; empty lines: limit ourselves to 2 newline
1322 ;; characters.
1330 ;;;; Plain List
1333 ;; Return structure of list at point. Internal function. See
1334 ;; `org-list-struct' for details.
1339 items struct)
1344 ;; At limit: end all items.
1350 ;; At list end: end all items.
1354 ;; At a new item: end previous sibling.
1366 ind
1367 bullet
1370 ;; Description tag.
1374 ;; Ending position, unknown so far.
1375 nil)))
1376 items))
1378 ;; Skip empty lines.
1380 ;; Skip inline tasks and blank lines along the way.
1387 ;; At some text line. Check if it ends any previous item.
1401 ;; Skip blocks (any type) and drawers contents.
1406 limit t)))
1412 "Parse a plain list.
1414 LIMIT bounds the search. AFFILIATED is a list of which CAR is
1415 the buffer position at the beginning of the first affiliated
1416 keyword and CDR is a plist of affiliated keywords along with
1417 their value. STRUCTURE is the structure of the plain list being
1418 parsed.
1420 Return a list whose CAR is `plain-list' and CDR is a plist
1421 containing `:type', `:begin', `:end', `:contents-begin' and
1422 `:contents-end', `:structure', `:post-blank' and
1423 `:post-affiliated' keywords.
1425 Assume point is at the beginning of the list."
1439 pos))
1443 ;; Return value.
1447 :begin begin
1448 :end end
1449 :contents-begin contents-begin
1450 :contents-end contents-end
1451 :structure struct
1457 "Interpret plain-list element as Org syntax.
1458 CONTENTS is the contents of the element."
1466 ;;;; Property Drawer
1469 "Parse a property drawer.
1471 LIMIT bounds the search.
1473 Return a list whose car is `property-drawer' and cdr is a plist
1474 containing `:begin', `:end', `:contents-begin', `:contents-end',
1475 `:post-blank' and `:post-affiliated' keywords.
1477 Assume point is at the beginning of the property drawer."
1490 :end end
1492 :contents-end contents-end
1497 "Interpret property-drawer element as Org syntax.
1498 CONTENTS is the properties within the drawer."
1502 ;;;; Quote Block
1505 "Parse a quote block.
1507 LIMIT bounds the search. AFFILIATED is a list of which CAR is
1508 the buffer position at the beginning of the first affiliated
1509 keyword and CDR is a plist of affiliated keywords along with
1510 their value.
1512 Return a list whose CAR is `quote-block' and CDR is a plist
1513 containing `:begin', `:end', `:contents-begin', `:contents-end',
1514 `:post-blank' and `:post-affiliated' keywords.
1516 Assume point is at the beginning of the block."
1520 ;; Incomplete block: parse it as a paragraph.
1526 ;; Empty blocks have no contents.
1539 :end end
1540 :contents-begin contents-begin
1541 :contents-end contents-end
1547 "Interpret quote-block element as Org syntax.
1548 CONTENTS is the contents of the element."
1552 ;;;; Section
1555 "Parse a section.
1557 Return a list whose CAR is `section' and CDR is a plist
1558 containing `:begin', `:end', `:contents-begin', `contents-end',
1559 `:post-blank' and `:post-affiliated' keywords."
1561 ;; Beginning of section is the beginning of the first non-blank
1562 ;; line after previous headline.
1570 :end end
1571 :contents-begin begin
1572 :contents-end pos-before-blank
1577 "Interpret section element as Org syntax.
1578 CONTENTS is the contents of the element."
1579 contents)
1582 ;;;; Special Block
1585 "Parse a special block.
1587 LIMIT bounds the search. AFFILIATED is a list of which CAR is
1588 the buffer position at the beginning of the first affiliated
1589 keyword and CDR is a plist of affiliated keywords along with
1590 their value.
1592 Return a list whose CAR is `special-block' and CDR is a plist
1593 containing `:type', `:begin', `:end', `:contents-begin',
1594 `:contents-end', `:post-blank' and `:post-affiliated' keywords.
1596 Assume point is at the beginning of the block."
1603 limit t)))
1604 ;; Incomplete block: parse it as a paragraph.
1610 ;; Empty blocks have no contents.
1623 :begin begin
1624 :end end
1625 :contents-begin contents-begin
1626 :contents-end contents-end
1632 "Interpret SPECIAL-BLOCK element as Org syntax.
1633 CONTENTS is the contents of the element."
1638 \f
1639 ;;; Elements
1640 ;;
1641 ;; For each element, a parser and an interpreter are also defined.
1642 ;; Both follow the same naming convention used for greater elements.
1643 ;;
1644 ;; Also, as for greater elements, adding a new element type is done
1645 ;; through the following steps: implement a parser and an interpreter,
1646 ;; tweak `org-element--current-element' so that it recognizes the new
1647 ;; type and add that new type to `org-element-all-elements'.
1650 ;;;; Babel Call
1653 "Parse a babel call.
1655 LIMIT bounds the search. AFFILIATED is a list of which car is
1656 the buffer position at the beginning of the first affiliated
1657 keyword and cdr is a plist of affiliated keywords along with
1658 their value.
1660 Return a list whose car is `babel-call' and cdr is a plist
1661 containing `:call', `:inside-header', `:arguments',
1662 `:end-header', `:begin', `:end', `:value', `:post-blank' and
1663 `:post-affiliated' as keywords."
1691 :inside-header inside-header
1692 :arguments arguments
1693 :end-header end-header
1694 :begin begin
1695 :end end
1696 :value value
1702 "Interpret BABEL-CALL element as Org syntax."
1712 ;;;; Clock
1715 "Parse a clock.
1717 LIMIT bounds the search.
1719 Return a list whose CAR is `clock' and CDR is a plist containing
1720 `:status', `:value', `:time', `:begin', `:end', `:post-blank' and
1721 `:post-affiliated' as keywords."
1741 :value value
1742 :duration duration
1743 :begin begin
1744 :end end
1745 :post-blank post-blank
1749 "Interpret CLOCK element as Org syntax."
1757 "%2s:%02s"
1761 ;;;; Comment
1764 "Parse a comment.
1766 LIMIT bounds the search. AFFILIATED is a list of which CAR is
1767 the buffer position at the beginning of the first affiliated
1768 keyword and CDR is a plist of affiliated keywords along with
1769 their value.
1771 Return a list whose CAR is `comment' and CDR is a plist
1772 containing `:begin', `:end', `:value', `:post-blank',
1773 `:post-affiliated' keywords.
1775 Assume point is at comment beginning."
1784 ;; Get comments ending.
1787 ;; Accumulate lines without leading hash and first
1788 ;; whitespace.
1802 :end end
1803 :value value
1809 "Interpret COMMENT element as Org syntax.
1810 CONTENTS is nil."
1814 ;;;; Comment Block
1817 "Parse an export block.
1819 LIMIT bounds the search. AFFILIATED is a list of which CAR is
1820 the buffer position at the beginning of the first affiliated
1821 keyword and CDR is a plist of affiliated keywords along with
1822 their value.
1824 Return a list whose CAR is `comment-block' and CDR is a plist
1825 containing `:begin', `:end', `:value', `:post-blank' and
1826 `:post-affiliated' keywords.
1828 Assume point is at comment block beginning."
1832 ;; Incomplete block: parse it as a paragraph.
1845 contents-begin contents-end)))
1849 :end end
1850 :value value
1856 "Interpret COMMENT-BLOCK element as Org syntax."
1863 ;;;; Diary Sexp
1866 "Parse a diary sexp.
1868 LIMIT bounds the search. AFFILIATED is a list of which CAR is
1869 the buffer position at the beginning of the first affiliated
1870 keyword and CDR is a plist of affiliated keywords along with
1871 their value.
1873 Return a list whose CAR is `diary-sexp' and CDR is a plist
1874 containing `:begin', `:end', `:value', `:post-blank' and
1875 `:post-affiliated' keywords."
1887 :begin begin
1888 :end end
1894 "Interpret DIARY-SEXP as Org syntax."
1898 ;;;; Example Block
1901 "Parse an example block.
1903 LIMIT bounds the search. AFFILIATED is a list of which CAR is
1904 the buffer position at the beginning of the first affiliated
1905 keyword and CDR is a plist of affiliated keywords along with
1906 their value.
1908 Return a list whose CAR is `example-block' and CDR is a plist
1909 containing `:begin', `:end', `:number-lines', `:preserve-indent',
1910 `:retain-labels', `:use-labels', `:label-fmt', `:switches',
1911 `:value', `:post-blank' and `:post-affiliated' keywords."
1915 ;; Incomplete block: parse it as a paragraph.
1923 ;; Switches analysis.
1927 switches)
1930 'new
1933 ;; Subtract 1 to give number of lines before
1934 ;; first line.
1938 ;; Should labels be retained in (or stripped from) example
1939 ;; blocks?
1944 ;; What should code-references use - labels or
1945 ;; line-numbers?
1954 ;; Standard block parsing.
1960 contents-begin contents-end)))
1969 :end end
1970 :value value
1971 :switches switches
1972 :number-lines number-lines
1973 :preserve-indent preserve-indent
1974 :retain-labels retain-labels
1975 :use-labels use-labels
1976 :label-fmt label-fmt
1982 "Interpret EXAMPLE-BLOCK element as Org syntax."
1989 val)
2002 ;;;; Export Block
2005 "Parse an export block.
2007 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2008 the buffer position at the beginning of the first affiliated
2009 keyword and CDR is a plist of affiliated keywords along with
2010 their value.
2012 Return a list whose CAR is `export-block' and CDR is a plist
2013 containing `:begin', `:end', `:type', `:value', `:post-blank' and
2014 `:post-affiliated' keywords.
2016 Assume point is at export-block beginning."
2020 ;; Incomplete block: parse it as a paragraph.
2039 contents-end))))
2043 :begin begin
2044 :end end
2045 :value value
2051 "Interpret EXPORT-BLOCK element as Org syntax."
2057 ;;;; Fixed-width
2060 "Parse a fixed-width section.
2062 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2063 the buffer position at the beginning of the first affiliated
2064 keyword and CDR is a plist of affiliated keywords along with
2065 their value.
2067 Return a list whose CAR is `fixed-width' and CDR is a plist
2068 containing `:begin', `:end', `:value', `:post-blank' and
2069 `:post-affiliated' keywords.
2071 Assume point is at the beginning of the fixed-width area."
2075 value
2080 ;; Accumulate text without starting colons.
2093 :end end
2094 :value value
2100 "Interpret FIXED-WIDTH element as Org syntax."
2108 ;;;; Horizontal Rule
2111 "Parse an horizontal rule.
2113 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2114 the buffer position at the beginning of the first affiliated
2115 keyword and CDR is a plist of affiliated keywords along with
2116 their value.
2118 Return a list whose CAR is `horizontal-rule' and CDR is a plist
2119 containing `:begin', `:end', `:post-blank' and `:post-affiliated'
2120 keywords."
2130 :end end
2136 "Interpret HORIZONTAL-RULE element as Org syntax."
2140 ;;;; Keyword
2143 "Parse a keyword at point.
2145 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2146 the buffer position at the beginning of the first affiliated
2147 keyword and CDR is a plist of affiliated keywords along with
2148 their value.
2150 Return a list whose CAR is `keyword' and CDR is a plist
2151 containing `:key', `:value', `:begin', `:end', `:post-blank' and
2152 `:post-affiliated' keywords."
2154 ;; An orphaned affiliated keyword is considered as a regular
2155 ;; keyword. In this case AFFILIATED is nil, so we take care of
2156 ;; this corner case.
2169 :value value
2170 :begin begin
2171 :end end
2177 "Interpret KEYWORD element as Org syntax."
2183 ;;;; Latex Environment
2187 "Regexp matching the beginning of a LaTeX environment.
2188 The environment is captured by the first group.
2194 "Format string matching the ending of a LaTeX environment.
2198 "Parse a LaTeX environment.
2200 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2201 the buffer position at the beginning of the first affiliated
2202 keyword and CDR is a plist of affiliated keywords along with
2203 their value.
2205 Return a list whose CAR is `latex-environment' and CDR is a plist
2206 containing `:begin', `:end', `:value', `:post-blank' and
2207 `:post-affiliated' keywords.
2209 Assume point is at the beginning of the latex environment."
2216 limit t))
2217 ;; Incomplete latex environment: parse it as a paragraph.
2227 :end end
2228 :value value
2234 "Interpret LATEX-ENVIRONMENT element as Org syntax."
2238 ;;;; Node Property
2241 "Parse a node-property at point.
2243 LIMIT bounds the search.
2245 Return a list whose CAR is `node-property' and CDR is a plist
2246 containing `:key', `:value', `:begin', `:end', `:post-blank' and
2247 `:post-affiliated' keywords."
2257 limit))))
2260 :value value
2261 :begin begin
2262 :end end
2267 "Interpret NODE-PROPERTY element as Org syntax."
2273 ;;;; Paragraph
2276 "Parse a paragraph.
2278 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2279 the buffer position at the beginning of the first affiliated
2280 keyword and CDR is a plist of affiliated keywords along with
2281 their value.
2283 Return a list whose CAR is `paragraph' and CDR is a plist
2284 containing `:begin', `:end', `:contents-begin' and
2285 `:contents-end', `:post-blank' and `:post-affiliated' keywords.
2287 Assume point is at the beginning of the paragraph."
2294 ;; A matching `org-element-paragraph-separate' is not
2295 ;; necessarily the end of the paragraph. In particular,
2296 ;; drawers, blocks or LaTeX environments opening lines
2297 ;; must be closed. Moreover keywords with a secondary
2298 ;; value must belong to "dual keywords".
2312 limit t)))
2318 limit t)))
2321 org-element-dual-keywords))
2322 ;; Everything else is unambiguous.
2335 :end end
2336 :contents-begin contents-begin
2337 :contents-end contents-end
2343 "Interpret paragraph element as Org syntax.
2344 CONTENTS is the contents of the element."
2345 contents)
2348 ;;;; Planning
2351 "Parse a planning.
2353 LIMIT bounds the search.
2355 Return a list whose CAR is `planning' and CDR is a plist
2356 containing `:closed', `:deadline', `:scheduled', `:begin',
2357 `:end', `:post-blank' and `:post-affiliated' keywords."
2367 closed deadline scheduled)
2379 :deadline deadline
2380 :scheduled scheduled
2381 :begin begin
2382 :end end
2383 :post-blank post-blank
2387 "Interpret PLANNING element as Org syntax."
2389 #'identity
2406 ;;;; Src Block
2409 "Parse a src block.
2411 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2412 the buffer position at the beginning of the first affiliated
2413 keyword and CDR is a plist of affiliated keywords along with
2414 their value.
2416 Return a list whose CAR is `src-block' and CDR is a plist
2417 containing `:language', `:switches', `:parameters', `:begin',
2418 `:end', `:number-lines', `:retain-labels', `:use-labels',
2419 `:label-fmt', `:preserve-indent', `:value', `:post-blank' and
2420 `:post-affiliated' keywords.
2422 Assume point is at the beginning of the block."
2425 limit t)))
2426 ;; Incomplete block: parse it as a paragraph.
2432 ;; Get language as a string.
2441 ;; Get switches.
2443 ;; Get parameters.
2445 ;; Switches analysis.
2449 switches)
2452 'new
2455 ;; Subtract 1 to give number of lines before
2456 ;; first line.
2464 ;; Should labels be retained in (or stripped from)
2465 ;; src blocks?
2470 ;; What should code-references use - labels or
2471 ;; line-numbers?
2476 ;; Retrieve code.
2483 ;; Get position after ending blank lines.
2493 :begin begin
2494 :end end
2495 :number-lines number-lines
2496 :preserve-indent preserve-indent
2497 :retain-labels retain-labels
2498 :use-labels use-labels
2499 :label-fmt label-fmt
2500 :value value
2506 "Interpret SRC-BLOCK element as Org syntax."
2515 val)
2530 ;;;; Table
2533 "Parse a table at point.
2535 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2536 the buffer position at the beginning of the first affiliated
2537 keyword and CDR is a plist of affiliated keywords along with
2538 their value.
2540 Return a list whose CAR is `table' and CDR is a plist containing
2541 `:begin', `:end', `:tblfm', `:type', `:contents-begin',
2542 `:contents-end', `:value', `:post-blank' and `:post-affiliated'
2543 keywords.
2545 Assume point is at the beginning of the table."
2561 acc))
2568 :end end
2569 :type type
2570 :tblfm tblfm
2571 ;; Only `org' tables have contents. `table.el' tables
2572 ;; use a `:value' property to store raw table as
2573 ;; a string.
2578 table-begin table-end))
2584 "Interpret TABLE element as Org syntax.
2585 CONTENTS is a string, if table's type is `org', or nil."
2596 ;;;; Table Row
2599 "Parse table row at point.
2601 Return a list whose CAR is `table-row' and CDR is a plist
2602 containing `:begin', `:end', `:contents-begin', `:contents-end',
2603 `:type', `:post-blank' and `:post-affiliated' keywords."
2607 ;; A table rule has no contents. In that case, ensure
2608 ;; CONTENTS-BEGIN matches CONTENTS-END.
2618 :begin begin
2619 :end end
2620 :contents-begin contents-begin
2621 :contents-end contents-end
2626 "Interpret TABLE-ROW element as Org syntax.
2627 CONTENTS is the contents of the table row."
2632 ;;;; Verse Block
2635 "Parse a verse block.
2637 LIMIT bounds the search. AFFILIATED is a list of which CAR is
2638 the buffer position at the beginning of the first affiliated
2639 keyword and CDR is a plist of affiliated keywords along with
2640 their value.
2642 Return a list whose CAR is `verse-block' and CDR is a plist
2643 containing `:begin', `:end', `:contents-begin', `:contents-end',
2644 `:post-blank' and `:post-affiliated' keywords.
2646 Assume point is at beginning of the block."
2650 ;; Incomplete block: parse it as a paragraph.
2665 :end end
2666 :contents-begin contents-begin
2667 :contents-end contents-end
2673 "Interpret verse-block element as Org syntax.
2674 CONTENTS is verse block contents."
2678 \f
2679 ;;; Objects
2680 ;;
2681 ;; Unlike to elements, raw text can be found between objects. Hence,
2682 ;; `org-element--object-lex' is provided to find the next object in
2683 ;; buffer.
2684 ;;
2685 ;; Some object types (e.g., `italic') are recursive. Restrictions on
2686 ;; object types they can contain will be specified in
2687 ;; `org-element-object-restrictions'.
2688 ;;
2689 ;; Creating a new type of object requires to alter
2690 ;; `org-element--object-regexp' and `org-element--object-lex', add the
2691 ;; new type in `org-element-all-objects', and possibly add
2692 ;; restrictions in `org-element-object-restrictions'.
2694 ;;;; Bold
2697 "Parse bold object at point, if any.
2699 When at a bold object, return a list whose car is `bold' and cdr
2700 is a plist with `:begin', `:end', `:contents-begin' and
2701 `:contents-end' and `:post-blank' keywords. Otherwise, return
2702 nil.
2704 Assume point is at the first star marker."
2716 :end end
2717 :contents-begin contents-begin
2718 :contents-end contents-end
2722 "Interpret bold object as Org syntax.
2723 CONTENTS is the contents of the object."
2727 ;;;; Code
2730 "Parse code object at point, if any.
2732 When at a code object, return a list whose car is `code' and cdr
2733 is a plist with `:value', `:begin', `:end' and `:post-blank'
2734 keywords. Otherwise, return nil.
2736 Assume point is at the first tilde marker."
2747 :begin begin
2748 :end end
2752 "Interpret CODE object as Org syntax."
2756 ;;;; Entity
2759 "Parse entity at point, if any.
2761 When at an entity, return a list whose car is `entity' and cdr
2762 a plist with `:begin', `:end', `:latex', `:latex-math-p',
2763 `:html', `:latin1', `:utf-8', `:ascii', `:use-brackets-p' and
2764 `:post-blank' as keywords. Otherwise, return nil.
2766 Assume point is at the beginning of the entity."
2768 (when (looking-at "\\\\\\(?:\\(?1:_ +\\)\\|\\(?1:there4\\|sup[123]\\|frac[13][24]\\|[a-zA-Z]+\\)\\(?2:$\\|{}\\|[^[:alpha:]]\\)\\)")
2786 :begin begin
2787 :end end
2788 :use-brackets-p bracketsp
2792 "Interpret ENTITY object as Org syntax."
2798 ;;;; Export Snippet
2801 "Parse export snippet at point.
2803 When at an export snippet, return a list whose car is
2804 `export-snippet' and cdr a plist with `:begin', `:end',
2805 `:back-end', `:value' and `:post-blank' as keywords. Otherwise,
2806 return nil.
2808 Assume point is at the beginning of the snippet."
2824 :value value
2825 :begin begin
2826 :end end
2830 "Interpret EXPORT-SNIPPET object as Org syntax."
2836 ;;;; Footnote Reference
2839 "Parse footnote reference at point, if any.
2841 When at a footnote reference, return a list whose car is
2842 `footnote-reference' and cdr a plist with `:label', `:type',
2843 `:begin', `:end', `:content-begin', `:contents-end' and
2844 `:post-blank' as keywords. Otherwise, return nil."
2860 :type type
2861 :begin begin
2862 :end end
2868 "Interpret FOOTNOTE-REFERENCE object as Org syntax.
2869 CONTENTS is its definition, when inline, or nil."
2875 ;;;; Inline Babel Call
2878 "Parse inline babel call at point, if any.
2880 When at an inline babel call, return a list whose car is
2881 `inline-babel-call' and cdr a plist with `:call',
2882 `:inside-header', `:arguments', `:end-header', `:begin', `:end',
2883 `:value' and `:post-blank' as keywords. Otherwise, return nil.
2885 Assume point is at the beginning of the babel call."
2899 ;; Parenthesis are mandatory.
2910 :inside-header inside-header
2911 :arguments arguments
2912 :end-header end-header
2913 :begin begin
2914 :end end
2915 :value value
2919 "Interpret INLINE-BABEL-CALL object as Org syntax."
2929 ;;;; Inline Src Block
2932 "Parse inline source block at point, if any.
2934 When at an inline source block, return a list whose car is
2935 `inline-src-block' and cdr a plist with `:begin', `:end',
2936 `:language', `:value', `:parameters' and `:post-blank' as
2937 keywords. Otherwise, return nil.
2939 Assume point is at the beginning of the inline src block."
2956 :value value
2957 :parameters parameters
2958 :begin begin
2963 "Interpret INLINE-SRC-BLOCK object as Org syntax."
2968 language
2970 body)))
2972 ;;;; Italic
2975 "Parse italic object at point, if any.
2977 When at an italic object, return a list whose car is `italic' and
2978 cdr is a plist with `:begin', `:end', `:contents-begin' and
2979 `:contents-end' and `:post-blank' keywords. Otherwise, return
2980 nil.
2982 Assume point is at the first slash marker."
2994 :end end
2995 :contents-begin contents-begin
2996 :contents-end contents-end
3000 "Interpret italic object as Org syntax.
3001 CONTENTS is the contents of the object."
3005 ;;;; Latex Fragment
3008 "Parse LaTeX fragment at point, if any.
3010 When at a LaTeX fragment, return a list whose car is
3011 `latex-fragment' and cdr a plist with `:value', `:begin', `:end',
3012 and `:post-blank' as keywords. Otherwise, return nil.
3014 Assume point is at the beginning of the LaTeX fragment."
3025 ;; Macro.
3048 :begin begin
3049 :end end
3053 "Interpret LATEX-FRAGMENT object as Org syntax."
3056 ;;;; Line Break
3059 "Parse line break at point, if any.
3061 When at a line break, return a list whose car is `line-break',
3062 and cdr a plist with `:begin', `:end' and `:post-blank' keywords.
3063 Otherwise, return nil.
3065 Assume point is at the beginning of the line break."
3074 "Interpret LINE-BREAK object as Org syntax."
3078 ;;;; Link
3081 "Parse link at point, if any.
3083 When at a link, return a list whose car is `link' and cdr a plist
3084 with `:type', `:path', `:format', `:raw-link', `:application',
3085 `:search-option', `:begin', `:end', `:contents-begin',
3086 `:contents-end' and `:post-blank' as keywords. Otherwise, return
3087 nil.
3089 Assume point is at the beginning of the link."
3092 end contents-begin contents-end link-end post-blank path type format
3093 raw-link search-option application)
3095 ;; Type 1: Text targeted from a radio target.
3105 ;; Type 2: Standard link, i.e. [[https://orgmode.org][homepage]]
3111 ;; RAW-LINK is the original link. Expand any
3112 ;; abbreviation in it.
3113 ;;
3114 ;; Also treat any newline character and associated
3115 ;; indentation as a single space character. This is not
3116 ;; compatible with RFC 3986, which requires to ignore
3117 ;; them altogether. However, doing so would require
3118 ;; users to encode spaces on the fly when writing links
3119 ;; (e.g., insert [[shell:ls%20*.org]] instead of
3120 ;; [[shell:ls *.org]], which defeats Org's focus on
3121 ;; simplicity.
3126 ;; Determine TYPE of link and set PATH accordingly. According
3127 ;; to RFC 3986, remove whitespaces from URI in external links.
3128 ;; In internal ones, treat indentation as a single space.
3130 ;; File type.
3135 ;; Explicit type (http, irc, bbdb...).
3139 ;; Code-ref type: PATH is the name of the reference.
3144 ;; Custom-id type: PATH is the name of the custom id.
3148 ;; Fuzzy type: Internal link either matches a target, an
3149 ;; headline name or nothing. PATH is the target or
3150 ;; headline's name.
3154 ;; Type 3: Plain link, e.g., https://orgmode.org
3161 ;; Type 4: Angular link, e.g., <https://orgmode.org>. Unlike to
3162 ;; bracket links, follow RFC 3986 and remove any extra
3163 ;; whitespace in URI.
3174 ;; In any case, deduce end point after trailing white space from
3175 ;; LINK-END variable.
3180 ;; Special "file" type link processing. Extract opening
3181 ;; application and search option, if any. Also normalize URI.
3188 ;; Translate link, if `org-link-translation-function' is set.
3196 :path path
3197 :format format
3199 :application application
3200 :search-option search-option
3201 :begin begin
3202 :end end
3203 :contents-begin contents-begin
3204 :contents-end contents-end
3208 "Interpret LINK object as Org syntax.
3209 CONTENTS is the contents of the object, or nil."
3214 ;; Links with contents and internal links have to
3215 ;; use bracket syntax. Ignore `:format' in these
3216 ;; cases. This is also the default syntax when the
3217 ;; property is not defined, e.g., when the object
3218 ;; was crafted by the user.
3221 ;; Since this is going to be used as
3222 ;; a format string, escape percent signs
3223 ;; in description.
3226 `nil
3229 ;; Otherwise, just obey to `:format'.
3241 path
3247 ;;;; Macro
3250 "Parse macro at point, if any.
3252 When at a macro, return a list whose car is `macro' and cdr
3253 a plist with `:key', `:args', `:begin', `:end', `:value' and
3254 `:post-blank' as keywords. Otherwise, return nil.
3256 Assume point is at the macro."
3272 :value value
3273 :args args
3274 :begin begin
3275 :end end
3279 "Interpret MACRO object as Org syntax."
3287 ;;;; Radio-target
3290 "Parse radio target at point, if any.
3292 When at a radio target, return a list whose car is `radio-target'
3293 and cdr a plist with `:begin', `:end', `:contents-begin',
3294 `:contents-end', `:value' and `:post-blank' as keywords.
3295 Otherwise, return nil.
3297 Assume point is at the radio target."
3309 :end end
3310 :contents-begin contents-begin
3311 :contents-end contents-end
3312 :post-blank post-blank
3316 "Interpret target object as Org syntax.
3317 CONTENTS is the contents of the object."
3321 ;;;; Statistics Cookie
3324 "Parse statistics cookie at point, if any.
3326 When at a statistics cookie, return a list whose car is
3327 `statistics-cookie', and cdr a plist with `:begin', `:end',
3328 `:value' and `:post-blank' keywords. Otherwise, return nil.
3330 Assume point is at the beginning of the statistics-cookie."
3341 :end end
3342 :value value
3346 "Interpret STATISTICS-COOKIE object as Org syntax."
3350 ;;;; Strike-Through
3353 "Parse strike-through object at point, if any.
3355 When at a strike-through object, return a list whose car is
3356 `strike-through' and cdr is a plist with `:begin', `:end',
3357 `:contents-begin' and `:contents-end' and `:post-blank' keywords.
3358 Otherwise, return nil.
3360 Assume point is at the first plus sign marker."
3372 :end end
3373 :contents-begin contents-begin
3374 :contents-end contents-end
3378 "Interpret strike-through object as Org syntax.
3379 CONTENTS is the contents of the object."
3383 ;;;; Subscript
3386 "Parse subscript at point, if any.
3388 When at a subscript object, return a list whose car is
3389 `subscript' and cdr a plist with `:begin', `:end',
3390 `:contents-begin', `:contents-end', `:use-brackets-p' and
3391 `:post-blank' as keywords. Otherwise, return nil.
3393 Assume point is at the underscore."
3407 :end end
3408 :use-brackets-p bracketsp
3409 :contents-begin contents-begin
3410 :contents-end contents-end
3414 "Interpret SUBSCRIPT object as Org syntax.
3415 CONTENTS is the contents of the object."
3418 contents))
3421 ;;;; Superscript
3424 "Parse superscript at point, if any.
3426 When at a superscript object, return a list whose car is
3427 `superscript' and cdr a plist with `:begin', `:end',
3428 `:contents-begin', `:contents-end', `:use-brackets-p' and
3429 `:post-blank' as keywords. Otherwise, return nil.
3431 Assume point is at the caret."
3445 :end end
3446 :use-brackets-p bracketsp
3447 :contents-begin contents-begin
3448 :contents-end contents-end
3452 "Interpret SUPERSCRIPT object as Org syntax.
3453 CONTENTS is the contents of the object."
3456 contents))
3459 ;;;; Table Cell
3462 "Parse table cell at point.
3463 Return a list whose car is `table-cell' and cdr is a plist
3464 containing `:begin', `:end', `:contents-begin', `:contents-end'
3465 and `:post-blank' keywords."
3473 :end end
3474 :contents-begin contents-begin
3475 :contents-end contents-end
3479 "Interpret table-cell element as Org syntax.
3480 CONTENTS is the contents of the cell, or nil."
3484 ;;;; Target
3487 "Parse target at point, if any.
3489 When at a target, return a list whose car is `target' and cdr
3490 a plist with `:begin', `:end', `:value' and `:post-blank' as
3491 keywords. Otherwise, return nil.
3493 Assume point is at the target."
3503 :end end
3504 :value value
3508 "Interpret TARGET object as Org syntax."
3512 ;;;; Timestamp
3523 "Parse time stamp at point, if any.
3525 When at a time stamp, return a list whose car is `timestamp', and
3526 cdr a plist with `:type', `:raw-value', `:year-start',
3527 `:month-start', `:day-start', `:hour-start', `:minute-start',
3528 `:year-end', `:month-end', `:day-end', `:hour-end',
3529 `:minute-end', `:repeater-type', `:repeater-value',
3530 `:repeater-unit', `:warning-type', `:warning-value',
3531 `:warning-unit', `:begin', `:end' and `:post-blank' keywords.
3532 Otherwise, return nil.
3534 Assume point is at the beginning of the timestamp."
3553 date-start)
3564 raw-value)
3566 :repeater-type
3572 :repeater-unit
3581 :warning-unit
3584 year-start month-start day-start hour-start minute-start year-end
3585 month-end day-end hour-end minute-end)
3586 ;; Parse date-start.
3594 ;; Compute date-end. It can be provided directly in time-stamp,
3595 ;; or extracted from time range. Otherwise, it defaults to the
3596 ;; same values as date-start.
3606 :raw-value raw-value
3607 :year-start year-start
3608 :month-start month-start
3609 :day-start day-start
3610 :hour-start hour-start
3611 :minute-start minute-start
3612 :year-end year-end
3613 :month-end month-end
3614 :day-end day-end
3615 :hour-end hour-end
3616 :minute-end minute-end
3617 :begin begin
3618 :end end
3620 repeater-props
3621 warning-props))))))
3624 "Interpret TIMESTAMP object as Org syntax."
3642 ;; Build an Org timestamp string from TIME. ACTIVEP is
3643 ;; non-nil when time stamp is active. If WITH-TIME-P is
3644 ;; non-nil, add a time part. HOUR-END and MINUTE-END
3645 ;; specify a time range in the timestamp. REPEAT-STRING is
3646 ;; the repeater string, if any.
3650 org-time-stamp-formats)
3651 time)))
3657 nil nil ts)))
3662 " "
3663 s
3665 ;; Return value.
3666 ts)))
3678 build-ts-string
3697 0
3705 "--"
3718 ;;;; Underline
3721 "Parse underline object at point, if any.
3723 When at an underline object, return a list whose car is
3724 `underline' and cdr is a plist with `:begin', `:end',
3725 `:contents-begin' and `:contents-end' and `:post-blank' keywords.
3726 Otherwise, return nil.
3728 Assume point is at the first underscore marker."
3740 :end end
3741 :contents-begin contents-begin
3742 :contents-end contents-end
3746 "Interpret underline object as Org syntax.
3747 CONTENTS is the contents of the object."
3751 ;;;; Verbatim
3754 "Parse verbatim object at point, if any.
3756 When at a verbatim object, return a list whose car is `verbatim'
3757 and cdr is a plist with `:value', `:begin', `:end' and
3758 `:post-blank' keywords. Otherwise, return nil.
3760 Assume point is at the first equal sign marker."
3771 :begin begin
3772 :end end
3776 "Interpret VERBATIM object as Org syntax."
3780 \f
3781 ;;; Parsing Element Starting At Point
3782 ;;
3783 ;; `org-element--current-element' is the core function of this section.
3784 ;; It returns the Lisp representation of the element starting at
3785 ;; point.
3786 ;;
3787 ;; `org-element--current-element' makes use of special modes. They
3788 ;; are activated for fixed element chaining (e.g., `plain-list' >
3789 ;; `item') or fixed conditional element chaining (e.g., `headline' >
3790 ;; `section'). Special modes are: `first-section', `item',
3791 ;; `node-property', `section' and `table-row'.
3794 "Parse the element starting at point.
3796 Return value is a list like (TYPE PROPS) where TYPE is the type
3797 of the element and PROPS a plist of properties associated to the
3798 element.
3800 Possible types are defined in `org-element-all-elements'.
3802 LIMIT bounds the search.
3804 Optional argument GRANULARITY determines the depth of the
3805 recursion. Allowed values are `headline', `greater-element',
3806 `element', `object' or nil. When it is broader than `object' (or
3807 nil), secondary values will not be parsed, since they only
3808 contain objects.
3810 Optional argument MODE, when non-nil, can be either
3811 `first-section', `section', `planning', `item', `node-property'
3812 and `table-row'.
3814 If STRUCTURE isn't provided but MODE is set to `item', it will be
3815 computed.
3817 This function assumes point is always at the beginning of the
3818 element it has to parse."
3821 ;; Determine if parsing depth allows for secondary strings
3822 ;; parsing. It only applies to elements referenced in
3823 ;; `org-element-secondary-value-alist'.
3826 ;; Item.
3829 ;; Table Row.
3831 ;; Node Property.
3833 ;; Headline.
3836 ;; Sections (must be checked after headline).
3841 limit)))
3842 ;; Planning.
3847 ;; Property drawer.
3853 ;; When not at bol, point is at the beginning of an item or
3854 ;; a footnote definition: next item is always a paragraph.
3856 ;; Clock.
3858 ;; Inlinetask.
3861 ;; From there, elements can have affiliated keywords.
3864 ;; Jumping over affiliated keywords put point off-limits.
3865 ;; Parse them as regular keywords.
3869 ;; LaTeX Environment.
3872 ;; Drawer and Property Drawer.
3875 ;; Fixed Width
3878 ;; Inline Comments, Blocks, Babel Calls, Dynamic Blocks and
3879 ;; Keywords.
3897 limit
3898 affiliated))
3911 ;; Footnote Definition.
3914 ;; Horizontal Rule.
3917 ;; Diary Sexp.
3920 ;; Table.
3923 ;; List.
3926 limit affiliated
3928 ;; Default element: Paragraph.
3932 ;; Most elements can have affiliated keywords. When looking for an
3933 ;; element beginning, we want to move before them, as they belong to
3934 ;; that element, and, in the meantime, collect information they give
3935 ;; into appropriate properties. Hence the following function.
3938 "Collect affiliated keywords from point down to LIMIT.
3940 Return a list whose CAR is the position at the first of them and
3941 CDR a plist of keywords and values and move point to the
3942 beginning of the first line after them.
3944 As a special case, if element doesn't start at the beginning of
3945 the line (e.g., a paragraph starting an item), CAR is current
3946 position of point and CDR is nil."
3950 ;; RESTRICT is the list of objects allowed in parsed
3951 ;; keywords value.
3953 output)
3956 ;; Apply translation to RAW-KWD. From there, KWD is
3957 ;; the official keyword.
3959 org-element-keyword-translation-alist))
3960 raw-kwd))
3961 ;; Find main value for any keyword.
3967 ;; PARSEDP is non-nil when keyword should have its
3968 ;; value parsed.
3970 ;; If KWD is a dual keyword, find its secondary
3971 ;; value. Maybe parse it.
3980 ;; Attribute a property name to KWD.
3982 ;; Now set final shape for VALUE.
3988 nil restrict)))
3992 ;; Attributes can always appear on multiple lines.
3995 ;; Eventually store the new value in OUTPUT.
3997 ;; Move to next keyword.
3999 ;; If affiliated keywords are orphaned: move back to first one.
4000 ;; They will be parsed as a paragraph.
4002 ;; Return value.
4006 \f
4007 ;;; The Org Parser
4008 ;;
4009 ;; The two major functions here are `org-element-parse-buffer', which
4010 ;; parses Org syntax inside the current buffer, taking into account
4011 ;; region, narrowing, or even visibility if specified, and
4012 ;; `org-element-parse-secondary-string', which parses objects within
4013 ;; a given string.
4014 ;;
4015 ;; The (almost) almighty `org-element-map' allows applying a function
4016 ;; on elements or objects matching some type, and accumulating the
4017 ;; resulting values. In an export situation, it also skips unneeded
4018 ;; parts of the parse tree.
4021 "Recursively parse the buffer and return structure.
4022 If narrowing is in effect, only parse the visible part of the
4023 buffer.
4025 Optional argument GRANULARITY determines the depth of the
4026 recursion. It can be set to the following symbols:
4028 `headline' Only parse headlines.
4029 `greater-element' Don't recurse into greater elements except
4030 headlines and sections. Thus, elements
4031 parsed are the top-level ones.
4032 `element' Parse everything but objects and plain text.
4033 `object' Parse the complete buffer (default).
4035 When VISIBLE-ONLY is non-nil, don't parse contents of hidden
4036 elements.
4038 An element or object is represented as a list with the
4039 pattern (TYPE PROPERTIES CONTENTS), where :
4041 TYPE is a symbol describing the element or object. See
4042 `org-element-all-elements' and `org-element-all-objects' for an
4043 exhaustive list of such symbols. One can retrieve it with
4044 `org-element-type' function.
4046 PROPERTIES is the list of attributes attached to the element or
4047 object, as a plist. Although most of them are specific to the
4048 element or object type, all types share `:begin', `:end',
4049 `:post-blank' and `:parent' properties, which respectively
4050 refer to buffer position where the element or object starts,
4051 ends, the number of white spaces or blank lines after it, and
4052 the element or object containing it. Properties values can be
4053 obtained by using `org-element-property' function.
4055 CONTENTS is a list of elements, objects or raw strings
4056 contained in the current element or object, when applicable.
4057 One can access them with `org-element-contents' function.
4059 The Org buffer has `org-data' as type and nil as properties.
4060 `org-element-map' function can be used to find specific elements
4061 or objects within the parse tree.
4063 This function assumes that current major mode is `org-mode'."
4069 ;; Start in `first-section' mode so text before the first
4070 ;; headline belongs to a section.
4074 "Recursively parse objects in STRING and return structure.
4076 RESTRICTION is a symbol limiting the object types that will be
4077 looked after.
4079 Optional argument PARENT, when non-nil, is the element or object
4080 containing the secondary string. It is used to set correctly
4081 `:parent' property within the string.
4083 If STRING is the empty string or nil, return nil."
4100 "Map a function on selected elements or objects.
4102 DATA is a parse tree, an element, an object, a string, or a list
4103 of such constructs. TYPES is a symbol or list of symbols of
4104 elements or objects types (see `org-element-all-elements' and
4105 `org-element-all-objects' for a complete list of types). FUN is
4106 the function called on the matching element or object. It has to
4107 accept one argument: the element or object itself.
4109 When optional argument INFO is non-nil, it should be a plist
4110 holding export options. In that case, parts of the parse tree
4111 not exportable according to that property list will be skipped.
4113 When optional argument FIRST-MATCH is non-nil, stop at the first
4114 match for which FUN doesn't return nil, and return that value.
4116 Optional argument NO-RECURSION is a symbol or a list of symbols
4117 representing elements or objects types. `org-element-map' won't
4118 enter any recursive element or object whose type belongs to that
4119 list. Though, FUN can still be applied on them.
4121 When optional argument WITH-AFFILIATED is non-nil, FUN will also
4122 apply to matching objects within parsed affiliated keywords (see
4123 `org-element-parsed-keywords').
4125 Nil values returned from FUN do not appear in the results.
4128 Examples:
4129 ---------
4131 Assuming TREE is a variable containing an Org buffer parse tree,
4132 the following example will return a flat list of all `src-block'
4133 and `example-block' elements in it:
4137 The following snippet will find the first headline with a level
4141 (lambda (hl)
4142 (and (= (org-element-property :level hl) 1)
4144 (org-element-property :begin hl)))
4145 nil t)
4147 The next example will return a flat list of all `plain-list' type
4148 elements in TREE that are not a sub-list themselves:
4152 Eventually, this example will return a flat list of all `bold'
4153 type objects containing a `latex-snippet' type object, even
4154 looking into captions:
4157 (lambda (b)
4159 nil nil nil t)"
4160 ;; Ensure TYPES and NO-RECURSION are a list, even of one element.
4164 ;; Recursion depth is determined by --CATEGORY.
4171 ;; If one object is found, the function has
4172 ;; to recurse into every object.
4175 ;; If one regular element is found, the
4176 ;; function has to recurse, at least, into
4177 ;; every element it encounters.
4180 --acc)
4183 ;; Recursively walk DATA. INFO, if non-nil, is a plist
4184 ;; holding contextual information.
4188 ;; Ignored element in an export context.
4190 ;; List of elements or objects.
4192 ;; Unconditionally enter parse trees.
4196 ;; Check if TYPE is matching among TYPES. If so,
4197 ;; apply FUN to --DATA and accumulate return value
4198 ;; into --ACC (or exit if FIRST-MATCH is non-nil).
4204 ;; If --DATA has a secondary string that can contain
4205 ;; objects with their type among TYPES, look inside.
4208 org-element-secondary-value-alist)))
4210 ;; If --DATA has any parsed affiliated keywords and
4211 ;; WITH-AFFILIATED is non-nil, look for objects in
4212 ;; them.
4219 ;; Pay attention to the type of parsed
4220 ;; keyword. In particular, preserve order for
4221 ;; multiple keywords.
4234 ;; Determine if a recursion into --DATA is possible.
4236 ;; --TYPE is explicitly removed from recursion.
4238 ;; --DATA has no contents.
4240 ;; Looking for greater elements but --DATA is
4241 ;; simply an element or an object.
4244 ;; Looking for elements but --DATA is an object.
4247 ;; In any other case, map contents.
4251 ;; Return value in a proper order.
4255 ;; The following functions are internal parts of the parser.
4256 ;;
4257 ;; The first one, `org-element--parse-elements' acts at the element's
4258 ;; level.
4259 ;;
4260 ;; The second one, `org-element--parse-objects' applies on all objects
4261 ;; of a paragraph or a secondary string. It calls
4262 ;; `org-element--object-lex' to find the next object in the current
4263 ;; container.
4266 "Return next special mode according to TYPE, or nil.
4267 TYPE is a symbol representing the type of an element or object
4268 containing next element if PARENTP is non-nil, or before it
4269 otherwise. Modes can be either `first-section', `item',
4270 `node-property', `planning', `property-drawer', `section',
4271 `table-row' or nil."
4288 "Parse elements between BEG and END positions.
4290 MODE prioritizes some elements over the others. It can be set to
4291 `first-section', `section', `planning', `item', `node-property'
4292 or `table-row'.
4294 When value is `item', STRUCTURE will be used as the current list
4295 structure.
4297 GRANULARITY determines the depth of the recursion. See
4298 `org-element-parse-buffer' for more information.
4300 When VISIBLE-ONLY is non-nil, don't parse contents of hidden
4301 elements.
4303 Elements are accumulated into ACC."
4306 ;; Visible only: skip invisible parts at the beginning of the
4307 ;; element.
4310 ;; When parsing only headlines, skip any text before first one.
4315 ;; Find current element's type and parse it accordingly to
4316 ;; its category.
4318 end granularity mode structure))
4322 ;; Visible only: skip invisible parts between siblings.
4325 ;; Fill ELEMENT contents by side-effect.
4327 ;; If element has no contents, don't modify it.
4329 ;; Greater element: parse it between `contents-begin' and
4330 ;; `contents-end'. Make sure GRANULARITY allows the
4331 ;; recursion, or ELEMENT is a headline, in which case going
4332 ;; inside is mandatory, in order to get sub-level headings.
4340 ;; Possibly switch to a special mode.
4344 granularity visible-only element))
4345 ;; ELEMENT has contents. Parse objects inside, if
4346 ;; GRANULARITY allows it.
4352 ;; Update mode.
4354 ;; Return result.
4358 "Return next object in current buffer or nil.
4359 RESTRICTION is a list of object types, as symbols, that should be
4360 looked after. This function assumes that the buffer is narrowed
4361 to an appropriate container (e.g., a paragraph)."
4365 ;; Object regexp sometimes needs to have a peek at
4366 ;; a character ahead. Therefore, when there is a hard
4367 ;; limit, make it one more than the true beginning of the
4368 ;; radio target.
4375 nil)
4376 ;; Since we moved backward, we do not want to
4377 ;; match again an hypothetical 1-character long
4378 ;; radio link before us. Realizing that this can
4379 ;; only happen if such a radio link starts at
4380 ;; beginning of line, we prevent this here.
4386 found)
4452 ;; This is probably a plain link.
4462 "Parse objects between BEG and END and return recursive structure.
4464 Objects are accumulated in ACC. RESTRICTION is a list of object
4465 successors which are allowed in the current object.
4467 ACC becomes the parent for all parsed objects. However, if ACC
4468 is nil (i.e., a secondary string is being parsed) and optional
4469 argument PARENT is non-nil, use it as the parent for all objects.
4470 Eventually, if both ACC and PARENT are nil, the common parent is
4471 the list of objects itself."
4479 ;; Text before any object.
4484 contents))))
4485 ;; Object...
4490 ;; Fill contents of NEXT-OBJECT if possible.
4492 cont-beg
4494 next-object
4496 next-object)
4497 contents)
4499 ;; Text after last object.
4503 contents)))
4504 ;; Result. Set appropriate parent.
4512 \f
4513 ;;; Towards A Bijective Process
4514 ;;
4515 ;; The parse tree obtained with `org-element-parse-buffer' is really
4516 ;; a snapshot of the corresponding Org buffer. Therefore, it can be
4517 ;; interpreted and expanded into a string with canonical Org syntax.
4518 ;; Hence `org-element-interpret-data'.
4519 ;;
4520 ;; The function relies internally on
4521 ;; `org-element--interpret-affiliated-keywords'.
4523 ;;;###autoload
4525 "Interpret DATA as Org syntax.
4526 DATA is a parse tree, an element, an object or a secondary string
4527 to interpret. Return Org syntax as a string."
4531 ;; Find interpreter for current object or
4532 ;; element. If it doesn't exist (e.g. this is
4533 ;; a pseudo object or element), return contents,
4534 ;; if any.
4541 ;; Secondary string.
4544 data
4546 ;; Full Org document.
4551 ;; Plain text: return it.
4553 ;; Element or object without contents.
4556 ;; Element or object with contents.
4559 interpret
4560 data
4561 ;; Recursively interpret contents.
4566 data
4567 ;; Fix indentation of elements containing
4568 ;; objects. We ignore `table-row'
4569 ;; elements as they are one line long
4570 ;; anyway.
4572 data
4573 ;; When normalizing first paragraph of
4574 ;; an item or a footnote-definition,
4575 ;; ignore first line's indentation.
4584 ;; Build white spaces. If no `:post-blank' property
4585 ;; is specified, assume its value is 0.
4595 "Return ELEMENT's affiliated keywords as Org syntax.
4596 If there is no affiliated keyword, return the empty string."
4606 ": "
4609 value)
4617 ;; All attribute keywords can have multiple lines.
4623 ;; List all ELEMENT's properties matching an attribute line or an
4624 ;; affiliated keyword, but ignore translated keywords since they
4625 ;; cannot belong to the property list.
4632 org-element-keyword-translation-alist)))))
4633 collect prop)
4636 ;; Because interpretation of the parse tree must return the same
4637 ;; number of blank lines between elements and the same number of white
4638 ;; space after objects, some special care must be given to white
4639 ;; spaces.
4640 ;;
4641 ;; The first function, `org-element-normalize-string', ensures any
4642 ;; string different from the empty string will end with a single
4643 ;; newline character.
4644 ;;
4645 ;; The second function, `org-element-normalize-contents', removes
4646 ;; global indentation from the contents of the current element.
4649 "Ensure string S ends with a single newline character.
4651 If S isn't a string return it unchanged. If S is the empty
4652 string, return it. Otherwise, return a new string with a single
4653 newline character at its end."
4661 "Normalize plain text in ELEMENT's contents.
4663 ELEMENT must only contain plain text and objects.
4665 If optional argument IGNORE-FIRST is non-nil, ignore first line's
4666 indentation to compute maximal common indentation.
4668 Return the normalized element that is element with global
4669 indentation removed from its contents."
4671 ;; Return minimal common indentation within BLOB. This is
4672 ;; done by walking recursively BLOB and updating MIN-IND
4673 ;; along the way. FIRST-FLAG is non-nil when the next
4674 ;; object is expected to be a string that doesn't start
4675 ;; with a newline character. It happens for strings at
4676 ;; the beginnings of the contents or right after a line
4677 ;; break.
4683 ;; Objects cannot start with spaces: in this
4684 ;; case, indentation is 0.
4725 ;; Build ELEMENT back, replacing each string with the same
4726 ;; string minus common indentation.
4729 ;; Return DATUM with all its strings indentation
4730 ;; shortened from MIN-IND white spaces.
4750 org-element-recursive-objects)
4754 datum)))
4758 \f
4759 ;;; Cache
4760 ;;
4761 ;; Implement a caching mechanism for `org-element-at-point' and
4762 ;; `org-element-context', which see.
4763 ;;
4764 ;; A single public function is provided: `org-element-cache-reset'.
4765 ;;
4766 ;; Cache is enabled by default, but can be disabled globally with
4767 ;; `org-element-use-cache'. `org-element-cache-sync-idle-time',
4768 ;; org-element-cache-sync-duration' and `org-element-cache-sync-break'
4769 ;; can be tweaked to control caching behavior.
4770 ;;
4771 ;; Internally, parsed elements are stored in an AVL tree,
4772 ;; `org-element--cache'. This tree is updated lazily: whenever
4773 ;; a change happens to the buffer, a synchronization request is
4774 ;; registered in `org-element--cache-sync-requests' (see
4775 ;; `org-element--cache-submit-request'). During idle time, requests
4776 ;; are processed by `org-element--cache-sync'. Synchronization also
4777 ;; happens when an element is required from the cache. In this case,
4778 ;; the process stops as soon as the needed element is up-to-date.
4779 ;;
4780 ;; A synchronization request can only apply on a synchronized part of
4781 ;; the cache. Therefore, the cache is updated at least to the
4782 ;; location where the new request applies. Thus, requests are ordered
4783 ;; from left to right and all elements starting before the first
4784 ;; request are correct. This property is used by functions like
4785 ;; `org-element--cache-find' to retrieve elements in the part of the
4786 ;; cache that can be trusted.
4787 ;;
4788 ;; A request applies to every element, starting from its original
4789 ;; location (or key, see below). When a request is processed, it
4790 ;; moves forward and may collide the next one. In this case, both
4791 ;; requests are merged into a new one that starts from that element.
4792 ;; As a consequence, the whole synchronization complexity does not
4793 ;; depend on the number of pending requests, but on the number of
4794 ;; elements the very first request will be applied on.
4795 ;;
4796 ;; Elements cannot be accessed through their beginning position, which
4797 ;; may or may not be up-to-date. Instead, each element in the tree is
4798 ;; associated to a key, obtained with `org-element--cache-key'. This
4799 ;; mechanism is robust enough to preserve total order among elements
4800 ;; even when the tree is only partially synchronized.
4804 "Non-nil when Org parser should cache its results.
4806 WARNING: for the time being, using cache sometimes triggers
4807 freezes. Therefore, it is disabled by default. Activate it if
4814 "Maximum duration, as a time value, for a cache synchronization.
4815 If the synchronization is not over after this delay, the process
4816 pauses and resumes after `org-element-cache-sync-break'
4820 "Duration, as a time value, of the pause between synchronizations.
4824 ;;;; Data Structure
4827 "AVL tree used to cache elements.
4828 Each node of the tree contains an element. Comparison is done
4829 with `org-element--cache-compare'. This cache is used in
4833 "List of pending synchronization requests.
4835 A request is a vector with the following pattern:
4837 \[NEXT BEG END OFFSET PARENT PHASE]
4839 Processing a synchronization request consists of three phases:
4841 0. Delete modified elements,
4842 1. Fill missing area in cache,
4843 2. Shift positions and re-parent elements after the changes.
4845 During phase 0, NEXT is the key of the first element to be
4846 removed, BEG and END is buffer position delimiting the
4847 modifications. Elements starting between them (inclusive) are
4848 removed. So are elements whose parent is removed. PARENT, when
4849 non-nil, is the parent of the first element to be removed.
4851 During phase 1, NEXT is the key of the next known element in
4852 cache and BEG its beginning position. Parse buffer between that
4853 element and the one before it in order to determine the parent of
4854 the next element. Set PARENT to the element containing NEXT.
4856 During phase 2, NEXT is the key of the next element to shift in
4857 the parse tree. All elements starting from this one have their
4858 properties relatives to buffer positions shifted by integer
4859 OFFSET and, if they belong to element PARENT, are adopted by it.
4867 "Hash table used to store keys during synchronization.
4871 "Return a unique key for ELEMENT in cache tree.
4873 Keys are used to keep a total order among elements in the cache.
4874 Comparison is done with `org-element--cache-key-less-p'.
4876 When no synchronization is taking place, a key is simply the
4877 beginning position of the element, or that position plus one in
4878 the case of an first item (respectively row) in
4879 a list (respectively a table).
4881 During a synchronization, the key is the one the element had when
4882 the cache was synchronized for the last time. Elements added to
4883 cache during the synchronization get a new key generated with
4884 `org-element--cache-generate-key'.
4886 Such keys are stored in `org-element--cache-sync-keys'. The hash
4887 table is cleared once the synchronization is complete."
4890 ;; Increase beginning position of items (respectively
4891 ;; table rows) by one, so the first item can get
4892 ;; a different key from its parent list (respectively
4893 ;; table).
4896 begin)))
4899 key))))
4902 "Generate a key between LOWER and UPPER.
4904 LOWER and UPPER are integers or lists, possibly empty.
4906 If LOWER and UPPER are equals, return LOWER. Otherwise, return
4907 a unique key, as an integer or a list of integers, according to
4908 the following rules:
4910 - LOWER and UPPER are compared level-wise until values differ.
4912 - If, at a given level, LOWER and UPPER differ from more than
4913 2, the new key shares all the levels above with LOWER and
4914 gets a new level. Its value is the mean between LOWER and
4915 UPPER:
4917 (1 2) + (1 4) --> (1 3)
4919 - If LOWER has no value to compare with, it is assumed that its
4920 value is `most-negative-fixnum'. E.g.,
4922 (1 1) + (1 1 2)
4924 is equivalent to
4926 (1 1 m) + (1 1 2)
4928 where m is `most-negative-fixnum'. Likewise, if UPPER is
4929 short of levels, the current value is `most-positive-fixnum'.
4931 - If they differ from only one, the new key inherits from
4932 current LOWER level and fork it at the next level. E.g.,
4934 (2 1) + (3 3)
4936 is equivalent to
4938 (2 1) + (2 M)
4940 where M is `most-positive-fixnum'.
4942 - If the key is only one level long, it is returned as an
4943 integer:
4945 (1 2) + (3 2) --> 2
4947 When they are not equals, the function assumes that LOWER is
4948 lesser than UPPER, per `org-element--cache-key-less-p'."
4952 skip-upper key)
4963 ;; When at a given level, LOWER and UPPER differ from
4964 ;; 1, ignore UPPER altogether. Instead create a key
4965 ;; between LOWER and the greatest key with the same
4966 ;; prefix as LOWER so far.
4972 "Non-nil if key A is less than key B.
4973 A and B are either integers or lists of integers, as returned by
4974 `org-element--cache-key'."
4982 ;; If A is empty, either keys are equal (B is also empty) and
4983 ;; we return nil, or A is lesser than B (B is longer) and we
4984 ;; return a non-nil value.
4985 ;;
4986 ;; If A is not empty, B is necessarily empty and A is greater
4987 ;; than B (A is longer). Therefore, return nil.
4991 "Non-nil when element A is located before element B."
4996 "Return root value in cache.
4997 This function assumes `org-element--cache' is a valid AVL tree."
5001 ;;;; Tools
5004 "Non-nil when cache is active in current buffer."
5006 org-element--cache
5010 "Find element in cache starting at POS or before.
5012 POS refers to a buffer position.
5014 When optional argument SIDE is non-nil, the function checks for
5015 elements starting at or past POS instead. If SIDE is `both', the
5016 function returns a cons cell where car is the first element
5017 starting at or before POS and cdr the first element starting
5018 after POS.
5020 The function can only find elements in the synchronized part of
5021 the cache."
5025 lower upper)
5040 ;; We found an element in cache starting at POS. If `side'
5041 ;; is `both' we also want the next one in order to generate
5042 ;; a key in-between.
5043 ;;
5044 ;; If the element is the first row or item in a table or
5045 ;; a plain list, we always return the table or the plain
5046 ;; list.
5047 ;;
5048 ;; In any other case, we return the element found.
5057 lower parent
5058 upper parent)))))
5061 lower element
5062 upper element)))))
5069 "Store ELEMENT in current buffer's cache, if allowed."
5072 ;; During synchronization, first build an appropriate key for
5073 ;; the new element so `avl-tree-enter' can insert it at the
5074 ;; right spot in the cache.
5083 org-element--cache-sync-keys)))
5087 "Remove ELEMENT from cache.
5088 Assume ELEMENT belongs to cache and that a cache is active."
5092 ;;;; Synchronization
5095 "Set idle timer for cache synchronization in BUFFER."
5102 org-element-cache-sync-idle-time))
5103 nil
5104 #'org-element--cache-sync
5105 buffer)))
5108 "Non-nil when synchronization process should be interrupted.
5109 TIME-LIMIT is a time value or nil."
5115 "Shift ELEMENT properties relative to buffer positions by OFFSET.
5117 Properties containing buffer positions are `:begin', `:end',
5118 `:contents-begin', `:contents-end' and `:structure'. When
5119 optional argument PROPS is a list of keywords, only shift
5120 properties provided in that list.
5122 Properties are modified by side-effect."
5124 ;; Shift `:structure' property for the first plain list only: it
5125 ;; is the only one that really matters and it prevents from
5126 ;; shifting it more than once.
5140 "Synchronize cache with recent modification in BUFFER.
5142 When optional argument THRESHOLD is non-nil, do the
5143 synchronization for all elements starting before or at threshold,
5144 then exit. Otherwise, synchronize cache for as long as
5145 `org-element-cache-sync-duration' or until Emacs leaves idle
5146 state.
5148 FUTURE-CHANGE, when non-nil, is a buffer position where changes
5149 not registered yet in the cache are going to happen. It is used
5150 in `org-element--cache-submit-request', where cache is partially
5151 updated before current modification are actually submitted."
5162 request
5164 threshold
5167 org-element-cache-sync-duration))
5168 future-change)
5169 ;; Request processed. Merge current and next offsets and
5170 ;; transfer ending position.
5176 ;; If more requests are awaiting, set idle timer accordingly.
5177 ;; Otherwise, reset keys.
5184 "Process synchronization REQUEST for all entries before NEXT.
5186 REQUEST is a vector, built by `org-element--cache-submit-request'.
5188 NEXT is a cache key, as returned by `org-element--cache-key'.
5190 When non-nil, THRESHOLD is a buffer position. Synchronization
5191 stops as soon as a shifted element begins after it.
5193 When non-nil, TIME-LIMIT is a time value. Synchronization stops
5194 after this time or when Emacs exits idle state.
5196 When non-nil, FUTURE-CHANGE is a buffer position where changes
5197 not registered yet in the cache are going to happen. See
5198 `org-element--cache-submit-request' for more information.
5200 Throw `interrupt' if the process stops before completing the
5201 request."
5204 ;; Phase 0.
5205 ;;
5206 ;; Delete all elements starting after BEG, but not after buffer
5207 ;; position END or past element with key NEXT. Also delete
5208 ;; elements contained within a previously removed element
5209 ;; (stored in `last-container').
5210 ;;
5211 ;; At each iteration, we start again at tree root since
5212 ;; a deletion modifies structure of the balanced tree.
5217 ;; Find first element in cache with key BEG or after it.
5221 data data-key last-container)
5230 data-key key
5233 data-key key
5234 node nil)))))
5244 up)))
5247 end))
5254 ;; No element starting after modifications left in
5255 ;; cache: further processing is futile.
5258 ;; Phase 1.
5259 ;;
5260 ;; Phase 0 left a hole in the cache. Some elements after it
5261 ;; could have parents within. For example, in the following
5262 ;; buffer:
5263 ;;
5264 ;; - item
5265 ;;
5266 ;;
5267 ;; Paragraph1
5268 ;;
5269 ;; Paragraph2
5270 ;;
5271 ;; if we remove a blank line between "item" and "Paragraph1",
5272 ;; everything down to "Paragraph2" is removed from cache. But
5273 ;; the paragraph now belongs to the list, and its `:parent'
5274 ;; property no longer is accurate.
5275 ;;
5276 ;; Therefore we need to parse again elements in the hole, or at
5277 ;; least in its last section, so that we can re-parent
5278 ;; subsequent elements, during phase 2.
5279 ;;
5280 ;; Note that we only need to get the parent from the first
5281 ;; element in cache after the hole.
5282 ;;
5283 ;; When next key is lesser or equal to the current one, delegate
5284 ;; phase 1 processing to next request in order to preserve key
5285 ;; order among requests.
5293 ;; Next element will start at its beginning position plus
5294 ;; offset, since it hasn't been shifted yet. Therefore, LIMIT
5295 ;; contains the real beginning position of the first element to
5296 ;; shift and re-parent.
5300 ;; Changes are going to happen around this element and
5301 ;; they will trigger another phase 1 request. Skip the
5302 ;; current one.
5308 ;; Phase 2.
5309 ;;
5310 ;; Shift all elements starting from key START, but before NEXT, by
5311 ;; OFFSET, and re-parent them when appropriate.
5312 ;;
5313 ;; Elements are modified by side-effect so the tree structure
5314 ;; remains intact.
5315 ;;
5316 ;; Once THRESHOLD, if any, is reached, or once there is an input
5317 ;; pending, exit. Before leaving, the current synchronization
5318 ;; request is updated.
5325 exit-flag)
5326 ;; No re-parenting nor shifting planned: request is over.
5336 ;; We reached NEXT. Request is complete.
5338 ;; Handle interruption request. Update current request.
5343 ;; Shift element.
5347 ;; Update PARENT and re-parent DATA, only when
5348 ;; necessary. Propagate new structures for lists.
5362 ;; Cache is up-to-date past THRESHOLD. Request
5363 ;; interruption.
5368 ;; We reached end of tree: synchronization complete.
5369 t)))
5372 "Parse elements in current section, down to POS.
5374 Start parsing from the closest between the last known element in
5375 cache or headline above. Return the smallest element containing
5376 POS.
5378 When optional argument SYNCP is non-nil, return the parent of the
5379 element containing POS instead. In that case, it is also
5380 possible to provide TIME-LIMIT, which is a time value specifying
5381 when the parsing should stop. The function throws `interrupt' if
5382 the process stopped before finding the expected result."
5389 element next mode)
5391 ;; Nothing in cache before point: start parsing from first
5392 ;; element following headline above, or first element in
5393 ;; buffer.
5400 ;; Cache returned exact match: return it.
5403 ;; There's a headline between cached value and POS: cached
5404 ;; value is invalid. Start parsing from first element
5405 ;; following the headline.
5412 ;; Check if CACHED or any of its ancestors contain point.
5413 ;;
5414 ;; If there is such an element, we inspect it in order to know
5415 ;; if we return it or if we need to parse its contents.
5416 ;; Otherwise, we just start parsing from current location,
5417 ;; which is right after the top-most element containing
5418 ;; CACHED.
5419 ;;
5420 ;; As a special case, if POS is at the end of the buffer, we
5421 ;; want to return the innermost element ending there.
5422 ;;
5423 ;; Also, if we find an ancestor and discover that we need to
5424 ;; parse its contents, make sure we don't start from
5425 ;; `:contents-begin', as we would otherwise go past CACHED
5426 ;; again. Instead, in that situation, we will resume parsing
5427 ;; from NEXT, which is located after CACHED or its higher
5428 ;; ancestor not containing point.
5440 ;; Parse successively each element until we reach POS.
5453 end 'element mode
5460 ;; Skip any element ending before point. Also skip
5461 ;; element ending at point (unless it is also the end of
5462 ;; buffer) since we're sure that another element begins
5463 ;; after it.
5467 ;; A non-greater element contains point: return it.
5470 ;; Otherwise, we have to decide if ELEMENT really
5471 ;; contains POS. In that case we start parsing from
5472 ;; contents' beginning.
5473 ;;
5474 ;; If POS is at contents' beginning but it is also at
5475 ;; the beginning of the first item in a list or a table.
5476 ;; In that case, we need to create an anchor for that
5477 ;; list or table, so return it.
5478 ;;
5479 ;; Also, if POS is at the end of the buffer, no element
5480 ;; can start after it, but more than one may end there.
5481 ;; Arbitrarily, we choose to return the innermost of
5482 ;; such elements.
5495 parent element
5496 end cend))))
5497 ;; Otherwise, return ELEMENT as it is the smallest
5498 ;; element containing POS.
5503 ;;;; Staging Buffer Changes
5514 "Regexp matching a sensitive line, structure wise.
5515 A sensitive line is a headline, inlinetask, block, drawer, or
5516 latex-environment boundary. When such a line is modified,
5517 structure changes in the document may propagate in the whole
5521 "Non-nil when a sensitive line is about to be changed.
5525 "Request extension of area going to be modified if needed.
5526 BEG and END are the beginning and end of the range of changed
5527 text. See `before-change-functions' for more information."
5537 'headline
5540 org-element--cache-sensitive-re bottom t)))))))))
5543 "Update buffer modifications for current buffer.
5544 BEG and END are the beginning and end of the range of changed
5545 text, and the length in bytes of the pre-change text replaced by
5546 that range. See `after-change-functions' for more information."
5554 ;; Determine if modified area needs to be extended, according
5555 ;; to both previous and current state. We make a special
5556 ;; case for headline editing: if a headline is modified but
5557 ;; not removed, do not extend.
5560 (`headline
5566 org-element--cache-sensitive-re bottom t))))
5567 ;; Effectively extend modified area.
5575 ;; Store synchronization request.
5578 ;; Activate a timer to process the request during idle time.
5582 "Return first element to remove from cache.
5584 BEG and END are buffer positions delimiting buffer modifications.
5585 OFFSET is the size of the changes.
5587 Returned element is usually the first element in cache containing
5588 any position between BEG and END. As an exception, greater
5589 elements around the changes that are robust to contents
5590 modifications are preserved and updated according to the
5591 changes."
5601 special-block))
5602 ;; Drawers named "PROPERTIES" are probably
5603 ;; a properties drawer being edited. Force
5604 ;; parsing to check if editing is over.
5613 ;; UP is a robust greater element containing changes.
5614 ;; We only need to extend its ending boundaries.
5620 ;; We're at top level element containing ELEMENT: if it's
5621 ;; altered by buffer modifications, it is first element in
5622 ;; cache to be removed. Otherwise, that first element is the
5623 ;; following one.
5624 ;;
5625 ;; As a special case, do not remove BEFORE if it is a robust
5626 ;; container for current changes.
5628 before)))))
5631 "Submit a new cache synchronization request for current buffer.
5632 BEG and END are buffer positions delimiting the minimal area
5633 where cache data should be removed. OFFSET is the size of the
5634 change, as an integer."
5636 delete-to delete-from)
5641 ;; Current changes can be merged with first sync request: we
5642 ;; can save a partial cache synchronization.
5645 ;; If last change happened within area to be removed, extend
5646 ;; boundaries of robust parents, if any. Otherwise, find
5647 ;; first element to remove and update request accordingly.
5659 ;; Ensure cache is correct up to END. Also make sure that NEXT,
5660 ;; if any, is no longer a 0-phase request, thus ensuring that
5661 ;; phases are properly ordered. We need to provide OFFSET as
5662 ;; optional parameter since current modifications are not known
5663 ;; yet to the otherwise correct part of the cache (i.e, before
5664 ;; the first request).
5671 ;; When changes happen before the first known
5672 ;; element, re-parent and shift the rest of the
5673 ;; cache.
5675 ;; Otherwise, we find the first non robust
5676 ;; element containing END. All elements between
5677 ;; FIRST and this one are to be removed.
5688 element up))
5690 org-element--cache-sync-requests)
5691 ;; No element to remove. No need to re-parent either.
5692 ;; Simply shift additional elements, if any, by OFFSET.
5695 offset)))))))
5698 ;;;; Public Functions
5700 ;;;###autoload
5702 "Reset cache in current buffer.
5703 When optional argument ALL is non-nil, reset cache in all Org
5704 buffers."
5721 ;;;###autoload
5723 "Refresh cache at position POS."
5730 \f
5731 ;;; The Toolbox
5732 ;;
5733 ;; The first move is to implement a way to obtain the smallest element
5734 ;; containing point. This is the job of `org-element-at-point'. It
5735 ;; basically jumps back to the beginning of section containing point
5736 ;; and proceed, one element after the other, with
5737 ;; `org-element--current-element' until the container is found. Note:
5738 ;; When using `org-element-at-point', secondary values are never
5739 ;; parsed since the function focuses on elements, not on objects.
5740 ;;
5741 ;; At a deeper level, `org-element-context' lists all elements and
5742 ;; objects containing point.
5743 ;;
5744 ;; `org-element-nested-p' and `org-element-swap-A-B' may be used
5745 ;; internally by navigation and manipulation tools.
5748 ;;;###autoload
5750 "Determine closest element around point.
5752 Return value is a list like (TYPE PROPS) where TYPE is the type
5753 of the element and PROPS a plist of properties associated to the
5754 element.
5756 Possible types are defined in `org-element-all-elements'.
5757 Properties depend on element or object type, but always include
5758 `:begin', `:end', `:parent' and `:post-blank' properties.
5760 As a special case, if point is at the very beginning of the first
5761 item in a list or sub-list, returned element will be that list
5762 instead of the item. Likewise, if point is at the beginning of
5763 the first row of a table, returned element will be the table
5764 instead of the first row.
5766 When point is at the end of the buffer, return the innermost
5767 element ending there."
5773 ;; Within blank lines at the beginning of buffer, return nil.
5775 ;; Within blank lines right after a headline, return that
5776 ;; headline.
5780 ;; Otherwise parse until we find element containing ORIGIN.
5787 ;;;###autoload
5789 "Return smallest element or object around point.
5791 Return value is a list like (TYPE PROPS) where TYPE is the type
5792 of the element or object and PROPS a plist of properties
5793 associated to it.
5795 Possible types are defined in `org-element-all-elements' and
5796 `org-element-all-objects'. Properties depend on element or
5797 object type, but always include `:begin', `:end', `:parent' and
5798 `:post-blank'.
5800 As a special case, if point is right after an object and not at
5801 the beginning of any other object, return that object.
5803 Optional argument ELEMENT, when non-nil, is the closest element
5804 containing point, as returned by `org-element-at-point'.
5805 Providing it allows for quicker computation."
5812 ;; If point is inside an element containing objects or
5813 ;; a secondary string, narrow buffer to the container and
5814 ;; proceed with parsing. Otherwise, return ELEMENT.
5816 ;; At a parsed affiliated keyword, check if we're inside main
5817 ;; or dual value.
5823 org-element-parsed-keywords))
5832 ;; Also change type to retrieve correct restrictions.
5834 ;; At an item, objects can only be located within tag, if any.
5845 ;; At an headline or inlinetask, objects are in title.
5857 ;; At a paragraph, a table-row or a verse block, objects are
5858 ;; located within their contents.
5862 ;; CBEG is nil for table rules.
5871 last)
5876 ;; Process NEXT, if any, in order to know if we need to
5877 ;; skip it, return it or move into it.
5884 ;; Skip objects ending before point. Also skip
5885 ;; objects ending at point unless it is also the
5886 ;; end of buffer, since we want to return the
5887 ;; innermost object.
5890 ;; For convenience, when object ends at POS,
5891 ;; without any space, store it in LAST, as we
5892 ;; will return it if no object starts here.
5896 ;; If POS is within a container object, move into
5897 ;; that object.
5901 ;; At contents' end, if there is no
5902 ;; space before point, also move into
5903 ;; object, for consistency with
5904 ;; convenience feature above.
5913 ;; Otherwise, return NEXT.
5917 "List all ancestors of a given element or object.
5919 DATUM is an object or element.
5921 Return ancestors from the closest to the farthest. When optional
5922 argument TYPES is a list of symbols, return the first element or
5923 object in the lineage whose type belongs to that list instead.
5925 When optional argument WITH-SELF is non-nil, lineage includes
5926 DATUM itself as the first element, and TYPES, if provided, also
5927 apply to it.
5929 When DATUM is obtained through `org-element-context' or
5930 `org-element-at-point', only ancestors from its section can be
5931 found. There is no such limitation when DATUM belongs to a full
5932 parse tree."
5934 ancestors)
5941 "Non-nil when elements ELEM-A and ELEM-B are nested."
5950 "Swap elements ELEM-A and ELEM-B.
5951 Assume ELEM-B is after ELEM-A in the buffer. Leave point at the
5952 end of ELEM-A."
5954 ;; There are two special cases when an element doesn't start at bol:
5955 ;; the first paragraph in an item or in a footnote definition.
5957 ;; Only a paragraph without any affiliated keyword can be moved at
5958 ;; ELEM-A position in such a situation. Note that the case of
5959 ;; a footnote definition is impossible: it cannot contain two
5960 ;; paragraphs in a row because it cannot contain a blank line.
5966 ;; In a special situation, ELEM-A will have no indentation. We'll
5967 ;; give it ELEM-B's (which will in, in turn, have no indentation).
5981 ;; Store inner overlays responsible for visibility status.
5982 ;; We also need to store their boundaries as they will be
5983 ;; removed from buffer.
5998 ;; Get contents.
6006 ;; Restore ex ELEM-A overlays.
6013 ;; Restore ex ELEM-B overlays.
6021 ;; Local variables:
6022 ;; generated-autoload-file: "org-loaddefs.el"
6023 ;; End:
6025 ;;; org-element.el ends here