1 # -*- encoding: utf-8 -*-
4 # Copyright (C) 2002-2004 Jörg Lehmann <joergl@users.sourceforge.net>
5 # Copyright (C) 2003-2004 Michael Schindler <m-schindler@users.sourceforge.net>
6 # Copyright (C) 2002-2012 André Wobst <wobsta@users.sourceforge.net>
8 # This file is part of PyX (http://pyx.sourceforge.net/).
10 # PyX 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 2 of the License, or
13 # (at your option) any later version.
15 # PyX 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 PyX; if not, write to the Free Software
22 # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
24 import math
, re
, configparser
, struct
, warnings
30 def splitatvalue(value
, *splitpoints
):
32 while section
< len(splitpoints
) and splitpoints
[section
] < value
:
34 if len(splitpoints
) > 1:
39 return (section
, value
)
42 _mathglobals
= {"neg": lambda x
: -x
,
43 "abs": lambda x
: x
< 0 and -x
or x
,
44 "sgn": lambda x
: x
< 0 and -1 or 1,
54 "sind": lambda x
: math
.sin(math
.pi
/180*x
),
55 "cosd": lambda x
: math
.cos(math
.pi
/180*x
),
56 "tand": lambda x
: math
.tan(math
.pi
/180*x
),
57 "asind": lambda x
: 180/math
.pi
*math
.asin(x
),
58 "acosd": lambda x
: 180/math
.pi
*math
.acos(x
),
59 "atand": lambda x
: 180/math
.pi
*math
.atan(x
),
60 "norm": lambda x
, y
: math
.hypot(x
, y
),
61 "splitatvalue": splitatvalue
,
67 """graph data interface
69 Graph data consists of columns, where each column might be identified by a
70 string or an integer. Each row in the resulting table refers to a data
73 All methods except for the constructor should consider self and its
74 attributes to be readonly, since the data instance might be shared between
75 several graphs simultaneously.
77 The instance variable columns is a dictionary mapping column names to the
78 data of the column (i.e. to a list). Only static columns (known at
79 construction time) are contained in that dictionary. For data with numbered
80 columns the column data is also available via the list columndata.
81 Otherwise the columndata list should be missing and an access to a column
84 The names of all columns (static and dynamic) must be fixed at the constructor
85 and stated in the columnnames dictionary.
87 The instance variable title and defaultstyles contain the data title and
88 the default styles (a list of styles), respectively. If defaultstyles is None,
89 the data cannot be plotted without user provided styles.
92 def dynamiccolumns(self
, graph
, axisnames
):
93 """create and return dynamic columns data
95 Returns dynamic data matching the given axes (the axes range and other
96 data might be used). The return value is a dictionary similar to the
97 columns instance variable. However, the static and dynamic data does
98 not need to be correlated in any way, i.e. the number of data points in
99 self.columns might differ from the number of data points represented by
100 the return value of the dynamiccolumns method.
105 defaultsymbols
= [style
.symbol()]
106 defaultlines
= [style
.line()]
111 defaultstyles
= defaultsymbols
113 def __init__(self
, title
="user provided values", **columns
):
114 for i
, values
in enumerate(list(columns
.values())):
115 if i
and len(values
) != l
:
116 raise ValueError("different number of values")
119 self
.columns
= columns
120 self
.columnnames
= list(columns
.keys())
125 "Graph data from a list of points"
127 defaultstyles
= defaultsymbols
129 def __init__(self
, points
, title
="user provided points", addlinenumbers
=1, **columns
):
132 self
.columndata
= [[x
] for x
in points
[0]]
133 for point
in points
[1:]:
135 raise ValueError("different number of columns per point")
136 for i
, x
in enumerate(point
):
137 self
.columndata
[i
].append(x
)
138 for v
in list(columns
.values()):
139 if abs(v
) > l
or (not addlinenumbers
and abs(v
) == l
):
140 raise ValueError("column number bigger than number of columns")
142 self
.columndata
= [list(range(1, len(points
) + 1))] + self
.columndata
143 self
.columns
= dict([(key
, self
.columndata
[i
]) for key
, i
in list(columns
.items())])
145 self
.columns
= dict([(key
, []) for key
, i
in list(columns
.items())])
146 self
.columnnames
= list(self
.columns
.keys())
153 _columnintref
= re
.compile(r
"\$(-?\d+)", re
.IGNORECASE
)
156 "creates a new data set out of an existing data set"
158 def __init__(self
, data
, title
=_notitle
, context
={}, copy
=1,
159 replacedollar
=1, columncallback
="__column__", **columns
):
161 if title
is _notitle
:
162 items
= list(columns
.items())
163 items
.sort() # we want sorted items (otherwise they would be unpredictable scrambled)
164 self
.title
= "%s: %s" % (text
.escapestring(data
.title
or "unkown source"),
165 ", ".join(["%s=%s" % (text
.escapestring(key
),
166 text
.escapestring(str(value
)))
167 for key
, value
in items
]))
172 self
.defaultstyles
= self
.orgdata
.defaultstyles
174 # analyse the **columns argument
176 for columnname
, value
in list(columns
.items()):
177 # search in the columns dictionary
179 self
.columns
[columnname
] = self
.orgdata
.columns
[value
]
181 # search in the columndata list
183 self
.columns
[columnname
] = self
.orgdata
.columndata
[value
]
184 except (AttributeError, TypeError):
185 # value was not an valid column identifier
186 # i.e. take it as a mathematical expression
188 m
= _columnintref
.search(value
)
190 value
= "%s%s(%s)%s" % (value
[:m
.start()], columncallback
, m
.groups()[0], value
[m
.end():])
191 m
= _columnintref
.search(value
)
192 value
= value
.replace("$", columncallback
)
193 expression
= compile(value
.strip(), __file__
, "eval")
194 context
= context
.copy()
195 context
[columncallback
] = self
.columncallback
196 if self
.orgdata
.columns
:
197 key
, columndata
= list(self
.orgdata
.columns
.items())[0]
198 count
= len(columndata
)
199 elif self
.orgdata
.columndata
:
200 count
= len(self
.orgdata
.columndata
[0])
204 for i
in range(count
):
205 self
.columncallbackcount
= i
206 for key
, values
in list(self
.orgdata
.columns
.items()):
207 context
[key
] = values
[i
]
209 newdata
.append(eval(expression
, _mathglobals
, context
))
210 except (ArithmeticError, ValueError):
212 self
.columns
[columnname
] = newdata
215 # copy other, non-conflicting column names
216 for columnname
, columndata
in list(self
.orgdata
.columns
.items()):
217 if columnname
not in self
.columns
:
218 self
.columns
[columnname
] = columndata
220 self
.columnnames
= list(self
.columns
.keys())
222 def columncallback(self
, value
):
224 return self
.orgdata
.columndata
[value
][self
.columncallbackcount
]
226 return self
.orgdata
.columns
[value
][self
.columncallbackcount
]
233 defaultcommentpattern
= re
.compile(r
"(#+|!+|%+)\s*")
234 defaultstringpattern
= re
.compile(r
"\"(.*?
)\"(\s
+|$
)")
235 defaultcolumnpattern = re.compile(r"(.*?
)(\s
+|$
)")
237 def splitline(self, line, stringpattern, columnpattern, tofloat=1):
238 """returns a tuple created out of the string line
239 - matches stringpattern and columnpattern, adds the first group of that
240 match to the result and and removes those matches until the line is empty
241 - when stringpattern matched, the result is always kept as a string
242 - when columnpattern matched and tofloat is true, a conversion to a float
243 is tried; when this conversion fails, the string is kept"""
245 # try to gain speed by skip matching regular expressions
246 if line.find('"')!=-1 or \
247 stringpattern is not self.defaultstringpattern or \
248 columnpattern is not self.defaultcolumnpattern:
250 match = stringpattern.match(line)
252 result.append(match.groups()[0])
253 line = line[match.end():]
255 match = columnpattern.match(line)
258 result.append(float(match.groups()[0]))
259 except (TypeError, ValueError):
260 result.append(match.groups()[0])
262 result.append(match.groups()[0])
263 line = line[match.end():]
267 return list(map(float, line.split()))
268 except (TypeError, ValueError):
270 for r in line.split():
272 result.append(float(r))
273 except (TypeError, ValueError):
279 def getcachekey(self, *args):
280 return ":".join([str(x) for x in args])
282 def __init__(self, filename,
283 commentpattern=defaultcommentpattern,
284 stringpattern=defaultstringpattern,
285 columnpattern=defaultcolumnpattern,
286 skiphead=0, skiptail=0, every=1,
289 def readfile(file, title, self=self, commentpattern=commentpattern, stringpattern=stringpattern, columnpattern=columnpattern, skiphead=skiphead, skiptail=skiptail, every=every):
294 for line in file.readlines():
296 match = commentpattern.match(line)
298 if not len(columndata):
299 columns = self.splitline(line[match.end():], stringpattern, columnpattern, tofloat=0)
302 for value in self.splitline(line, stringpattern, columnpattern, tofloat=1):
303 linedata.append(value)
305 if linenumber >= skiphead and not ((linenumber - skiphead) % every):
306 linedata = [linenumber + 1] + linedata
307 if len(linedata) > maxcolumns:
308 maxcolumns = len(linedata)
309 columndata.append(linedata)
311 if skiptail >= every:
312 skip, x = divmod(skiptail, every)
313 del columndata[-skip:]
314 for i in range(len(columndata)):
315 if len(columndata[i]) != maxcolumns:
316 columndata[i].extend([None]*(maxcolumns-len(columndata[i])))
317 return points(columndata, title=title, addlinenumbers=0,
318 **dict([(column, i+1) for i, column in enumerate(columns[:maxcolumns-1])]))
323 # not a file-like object -> open it
324 cachekey = self.getcachekey(filename, commentpattern, stringpattern, columnpattern, skiphead, skiptail, every)
325 if cachekey not in filecache:
326 filecache[cachekey] = readfile(open(filename), filename)
327 data.__init__(self, filecache[cachekey], **kwargs)
329 data.__init__(self, readfile(filename, "user provided file-like object"), **kwargs)
334 class conffile(data):
336 def __init__(self, filename, **kwargs):
337 """read data from a config-like file
338 - filename is a string
339 - each row is defined by a section in the config-like file (see
340 config module description)
341 - the columns for each row are defined by lines in the section file;
342 the option entries identify and name the columns
343 - further keyword arguments are passed to the constructor of data,
344 keyword arguments data and titles excluded"""
346 def readfile(file, title):
347 config = configparser.ConfigParser(strict=False)
348 config.optionxform = str
349 config.read_file(file)
350 sections = config.sections()
352 columndata = [None]*len(sections)
355 for i in range(len(sections)):
356 point = [sections[i]] + [None]*(maxcolumns-1)
357 for option in config.options(sections[i]):
358 value = config.get(sections[i], option)
364 index = columns[option]
366 columns[option] = maxcolumns
371 columndata[i] = point
372 # wrap result into a data instance to remove column numbers
373 result = data(points(columndata, addlinenumbers=0, **columns), title=title)
374 # ... but reinsert sections as linenumbers
375 result.columndata = [[x[0] for x in columndata]]
381 # not a file-like object -> open it
382 if filename not in filecache:
383 filecache[filename] = readfile(open(filename), filename)
384 data.__init__(self, filecache[filename], **kwargs)
386 data.__init__(self, readfile(filename, "user provided file-like object"), **kwargs)
393 defaultstyles = defaultlines
395 def getcachekey(self, *args):
396 return ":".join([str(x) for x in args])
398 def __init__(self, filename, minrank=None, maxrank=None, **kwargs):
402 def __init__(self, file):
408 self.fill) = struct.unpack("<5i20s", file.read(40))
409 if self.magic != 0x20770002:
410 raise ValueError("bad magic number")
414 def __init__(self, file, i):
423 self.rank) = struct.unpack("<6i2h", file.read(28))
427 def __init__(self, file, sd):
428 file.seek(sd.absaddr)
433 self.dummy) = struct.unpack("<3i2h", file.read(16))
434 oln, olt = self.orgx, self.orgy
435 self.points = [(olt, oln)]
436 for i in range(self.nstrokes):
437 c1, c2 = struct.unpack("2c", file.read(2))
448 c3, c4, c5, c6, c7, c8 = struct.unpack("6c", file.read(6))
450 c2 = chr(ord(c2) | 0x40)
451 dx, dy = struct.unpack("<2i", c3+c4+c1+c2+c7+c8+c5+c6)
454 self.points.append((olt, oln))
455 sd.nstrokes = self.nstrokes
457 def readfile(file, title):
459 file.seek(h.dictaddr)
460 sds = [segdict(file, i+1) for i in range(h.segcount)]
461 sbs = [segment(file, sd) for sd in sds]
463 # remove jumps at long +/- 180
464 for sd, sb in zip(sds, sbs):
465 if sd.minlong < -150*3600 and sd.maxlong > 150*3600:
466 for i, (lat, int) in enumerate(sb.points):
468 sb.points[i] = lat, int + 360*3600
471 for sd, sb in zip(sds, sbs):
472 if ((minrank is None or sd.rank >= minrank) and
473 (maxrank is None or sd.rank <= maxrank)):
475 columndata.append((None, None))
476 columndata.extend([(int/3600.0, lat/3600.0)
477 for lat, int in sb.points])
479 result = points(columndata, title=title)
480 result.defaultstyles = self.defaultstyles
487 # not a file-like object -> open it
488 cachekey = self.getcachekey(filename, minrank, maxrank)
489 if cachekey not in cbdfilecache:
490 cbdfilecache[cachekey] = readfile(open(filename, "rb"), filename)
491 data.__init__(self, cbdfilecache[cachekey], **kwargs)
493 data.__init__(self, readfile(filename, "user provided file-like object"), **kwargs)
496 class function(_data):
498 defaultstyles = defaultlines
500 assignmentpattern = re.compile(r"\s*([a-z_][a-z0-9_]*)\s*\(\s*([a-z_][a-z0-9_]*)\s*\)\s*=", re.IGNORECASE)
502 def __init__(self, expression, title=_notitle, min=None, max=None,
503 points=100, context={}):
505 if title is _notitle:
506 self.title = expression
511 self.numberofpoints = points
512 self.context = context.copy() # be safe on late evaluations
513 m = self.assignmentpattern.match(expression)
515 self.yname, self.xname = m.groups()
516 expression = expression[m.end():]
518 raise ValueError("y(x)=... or similar expected")
519 if self.xname in context:
520 raise ValueError("xname in context")
521 self.expression = compile(expression.strip(), __file__, "eval")
523 self.columnnames = [self.xname, self.yname]
525 def dynamiccolumns(self, graph, axisnames):
526 dynamiccolumns = {self.xname: [], self.yname: []}
528 xaxis = graph.axes[axisnames.get(self.xname, self.xname)]
529 from pyx.graph.axis import logarithmic
530 logaxis = isinstance(xaxis.axis, logarithmic)
531 if self.min is not None:
535 if self.max is not None:
542 for i in range(self.numberofpoints):
543 x = min + (max-min)*i / (self.numberofpoints-1.0)
546 dynamiccolumns[self.xname].append(x)
547 self.context[self.xname] = x
549 y = eval(self.expression, _mathglobals, self.context)
550 except (ArithmeticError, ValueError):
552 dynamiccolumns[self.yname].append(y)
553 return dynamiccolumns
556 class functionxy(function):
558 def __init__(self, f, min=None, max=None, **kwargs):
559 function.__init__(self, "y(x)=f(x)", context={"f": f}, min=min, max=max, **kwargs)
562 class paramfunction(_data):
564 defaultstyles = defaultlines
566 def __init__(self, varname, min, max, expression, title=_notitle, points=100, context={}):
567 if varname in context:
568 raise ValueError("varname in context")
569 if title is _notitle:
570 self.title = expression
573 varlist, expression = expression.split("=")
574 expression = compile(expression.strip(), __file__, "eval")
575 keys = [key.strip() for key in varlist.split(",")]
576 self.columns = dict([(key, []) for key in keys])
577 context = context.copy()
578 for i in range(points):
579 param = min + (max-min)*i / (points-1.0)
580 context[varname] = param
581 values = eval(expression, _mathglobals, context)
582 for key, value in zip(keys, values):
583 self.columns[key].append(value)
584 if len(keys) != len(values):
585 raise ValueError("unpack tuple of wrong size")
586 self.columnnames = list(self.columns.keys())
589 class paramfunctionxy(paramfunction):
591 def __init__(self, f, min, max, **kwargs):
592 paramfunction.__init__(self, "t", min, max, "x, y = f(t)", context={"f": f}, **kwargs)
595 class _nodefaultstyles:
600 "creates a new data set by joining from a list of data, it does however *not* combine points, but fills data with None if necessary"
602 def merge_lists(self, lists):
603 "merges list items w/o duplications, resulting order is arbitraty"
606 result.update(set(l))
607 return builtinlist(result)
609 def merge_dicts(self, dicts):
610 """merge dicts containing lists as values (with equal number of items
611 per list in each dict), missing data is padded by None"""
612 keys = self.merge_lists([list(d.keys()) for d in dicts])
615 if len(list(d.keys())) == len(keys):
616 empties.append(None) # won't be needed later on
618 values
= list(d
.values())
620 empties
.append([None]*len(values
[0]))
622 # has no data at all -> do not add anything
627 for d
, e
in zip(dicts
, empties
):
628 result
[key
].extend(d
.get(key
, e
))
631 def __init__(self
, data
, title
=_notitle
, defaultstyles
=_nodefaultstyles
):
632 """takes a list of data, a title (if it should not be autoconstructed)
633 and a defaultstyles list if there is no common defaultstyles setting
634 for in the provided data"""
637 self
.columnnames
= self
.merge_lists([d
.columnnames
for d
in data
])
638 self
.columns
= self
.merge_dicts([d
.columns
for d
in data
])
639 if title
is _notitle
:
640 self
.title
= " + ".join([d
.title
for d
in data
])
643 if defaultstyles
is _nodefaultstyles
:
644 self
.defaultstyles
= data
[0].defaultstyles
646 if d
.defaultstyles
is not self
.defaultstyles
:
647 self
.defaultstyles
= None
650 self
.defaultstyles
= defaultstyles
652 def dynamiccolumns(self
, graph
, axisnames
):
653 return self
.merge_dicts([d
.dynamiccolumns(graph
, axisnames
) for d
in self
.data
])