Wahrscheinlichere Alternative an erster Stelle.
[wortliste.git] / Makefile
blobf89597a5943520cb423c6685c0031520dbbe729c
1 # -*- coding: utf-8 -*-
3 # This Makefile creates German hyphenation patterns in subdirectories
4 # $(TRAD) and $(REFO) for traditional and new orthography, respectively.
5 # Hyphenation patterns for traditional Swiss German are generated in
6 # directory $(SWISS).
8 # The input data is in $(SRCDIR); the possible targets are `pattern-trad',
9 # `pattern-refo', and `pattern-swiss'. If no target (or target `all') is
10 # given, all patterns for all three targets are built.
12 # SRCDIR (and the other variables) can be easily modified as parameters
13 # while calling `make', e.g.
15 # make pattern-trad SRCDIR=~/git/wortliste
17 # If you add one of the (phony) targets `major', `fugen', or `suffix',
18 # patterns that only use major hyphenation points (`Haupttrennstellen')
19 # are created. Example:
21 # make major pattern-refo
23 # The used directories names are the same as above but with `-major' (etc.)
24 # appended to the names.
26 # To control the used weights in the major hyphenation patterns, add
27 # variable `W=N', where `N' gives the quality: value 1 specifies the best
28 # hyphenation points only, value 2 both the best and second-best points,
29 # etc. The default is value 0, using all major hyphenation points.
33 # Dieses Makefile erzeugt deutsche Trennmuster in den
34 # Unterverzeichnissen $(TRAD) und $(REFO) für die traditionelle
35 # bzw. reformierte Rechtschreibung. Trennmuster für tradionelles
36 # deutschschweizerisches Deutsch werden Verzeichnis $(SWISS) erzeugt.
38 # Die Eingabedaten werden im Verzeichnis $(SRCDIR) erwartet; die möglichen
39 # Make-Ziele sind `pattern-trad', `pattern-refo' und `pattern-swiss'. Wenn
40 # kein Ziel angegeben ist (oder man das Ziel `all' verwendet), werden alle
41 # drei Trennmuster erzeugt.
43 # SRCDIR (und die anderen Variablen) kann man leicht beim Aufruf von
44 # `make' als Parameter modifizieren, z.B.
46 # make pattern-trad SRCDIR=~/git/wortliste
48 # Wird eines der zusätzlichen (künstlichen) Ziele `major', `fugen' oder
49 # `suffix' angegeben, werden Haupttrennstellmuster erzeugt.
51 # Beispiel:
53 # make major pattern-refo
55 # Die verwendeten Verzeichnisnamen sind die gleichen wie oben, allerdings
56 # mit einem angehängten `-major', `-fugen' bzw. `-suffix'.
58 # Diese Spezialmuster spiegeln die Auszeichnung in der Liste direkt wider.
59 # Sie haben nicht das Ziel, "gute" Trennungen in Texten zu erzeugen sondern
60 # sind zum Testen der Konsistenz der Auszeichnung sowie zum "kategorisierten"
61 # Markieren der Trennstellen neuer Wörter gedacht.
63 # Die Wichtung der verwendeten Haupttrennstellen kann mittels der Variable
64 # `W=N' kontrolliert werden, wo `N' die Qualität angibt: Wert 1 selektiert
65 # nur die besten Haupttrennstellen, Wert 2 die besten und zweitbesten
66 # Haupttrennstellen usw. Der Standardwert für `W' ist 0; er gibt an, dass
67 # alle Haupttrennstellen verwendet werden sollen.
70 SRCDIR = .
71 W = 0
73 DATADIR = $(SRCDIR)/daten
74 SCRIPTDIR = $(SRCDIR)/skripte
75 PYSCRIPTDIR = $(SRCDIR)/skripte/python
76 WORDLIST = wortliste
77 # Variables FROM and TO are used by goal `pidiff'. FROM must be a
78 # commit set from shell, like `make pidiff FROM=abcdef', TO is optional
79 # and evaluates to `HEAD' if not given.
81 ifneq ($(findstring major,$(MAKECMDGOALS)),)
82 MAJOR = -major
83 # A single `-' gets removed; all other combinations of `-', `<', `>',
84 # and `=' are converted to a hyphen.
85 SEDMAJOR = $(SED) -e '/[=<>-]/!n' \
86 -e 's/---*/=/g' \
87 -e 's/-//g' \
88 -e 's/[=<>][=<>]*/-/g' \
89 -e '/-/!d'
90 PERLMAJOR = -g $(W)
92 ifeq ($(words $(MAKECMDGOALS)),1)
93 major: all
94 else
95 # This is to suppress the `nothing to be done' warning.
96 major:
98 endif
99 else ifneq ($(findstring fugen,$(MAKECMDGOALS)),)
100 MAJOR = -fugen
101 # All combinations of `-', `<', `>', `<=', `=>' get removed,
102 # runs of `=' are converted to a hyphen.
103 SEDMAJOR = $(SED) -e '/[=<>-]/!n' \
104 -e 's/--*//g' \
105 -e 's/<=*//g' \
106 -e 's/=*>//g' \
107 -e 's/[<>][<>]*//g' \
108 -e 's/[=][=]*/-/g'
109 PERLMAJOR = -g $(W)
111 ifeq ($(words $(MAKECMDGOALS)),1)
112 fugen: all
113 else
114 # This is to suppress the `nothing to be done' warning.
115 fugen:
117 endif
118 else ifneq ($(findstring suffix,$(MAKECMDGOALS)),)
119 MAJOR = -suffix
120 # All combinations of `-', `<', `=' get removed,
121 # runs of `>' are converted to a hyphen.
122 SEDMAJOR = $(SED) -e '/[=<>-]/!n' \
123 -e 's/-//g' \
124 -e 's/[<=][<=]*//g' \
125 -e 's/[>][>]*/-/g'
126 PERLMAJOR = -g $(W)
128 ifeq ($(words $(MAKECMDGOALS)),1)
129 suffix: all
130 else
131 # This is to suppress the `nothing to be done' warning.
132 suffix:
134 endif
135 else
136 MAJOR =
137 SEDMAJOR = cat
138 PERLMAJOR =
139 endif
141 TRAD = dehypht-x$(MAJOR)
142 REFO = dehyphn-x$(MAJOR)
143 SWISS = dehyphts-x$(MAJOR)
144 LANGS = de-Latf
146 LC_ENVVARS = LC_COLLATE=de_DE.ISO8859-15 \
147 LC_CTYPE=de_DE.ISO8859-15
149 CAT = cat
150 CHDIR = cd
151 COPY = cp
152 DATE = $(shell date '+%Y-%m-%d')
153 ECHO = echo
154 GIT = git
155 ICONV = iconv -f iso-8859-15 -t utf-8
156 MKDIR = mkdir -p
157 PERL = perl
158 PWD = pwd
159 PYTHON = python
160 SED = sed
161 SH = bash
162 SORT = $(LC_ENVVARS) sort -d \
163 | $(LC_ENVVARS) uniq -i
165 GIT_VERSION := `$(CHDIR) $(SRCDIR); \
166 $(GIT) log --format=%H -1 HEAD --`
167 TRADFILES = $(TRAD)/$(TRAD)-$(DATE).pat $(TRAD)/$(TRAD)-$(DATE).tex
168 REFOFILES = $(REFO)/$(REFO)-$(DATE).pat $(REFO)/$(REFO)-$(DATE).tex
169 SWISSFILES = $(SWISS)/$(SWISS)-$(DATE).pat $(SWISS)/$(SWISS)-$(DATE).tex
172 override SRCDIR := $(shell cd $(SRCDIR); $(PWD))
175 all: pattern-trad pattern-refo pattern-swiss
177 .PHONY: pattern-trad pattern-refo pattern-swiss major fugen suffix
178 pattern-trad: $(TRADFILES)
179 pattern-refo: $(REFOFILES)
180 pattern-swiss: $(SWISSFILES)
182 # intermediate targets
184 # auxiliary targets
186 .PHONY: words-trad words-refo
187 words-trad: $(TRAD)/words.hyphenated.trad
188 words-refo: $(REFO)/words.hyphenated.refo
191 .PHONY: pre-trad pre-refo pre-swiss
192 pre-trad:
193 $(MKDIR) $(TRAD)
194 pre-refo:
195 $(MKDIR) $(REFO)
196 pre-swiss:
197 $(MKDIR) $(SWISS)
199 $(TRADFILES) $(TRAD)/words.hyphenated.trad: pre-trad
200 $(REFOFILES) $(REFO)/words.hyphenated.refo: pre-refo
201 $(SWISSFILES) $(SWISS)/words.hyphenated.swiss: pre-swiss
204 # GNU make supports creation of multiple targets by a single
205 # invocation of a recipe only for pattern rules, thus we have
206 # to use a `sentinel file' (using `echo' for the time stamp).
209 $(TRAD)/pattern.8 $(TRAD)/pattern.rules: $(TRAD)/make-full-pattern-trad
211 $(TRAD)/make-full-pattern-trad: $(TRAD)/words.hyphenated.trad
212 $(CHDIR) $(TRAD); \
213 $(SH) $(SCRIPTDIR)/make-full-pattern.sh $(<F) $(DATADIR)/german.tr
214 $(ECHO) done > $@
216 $(TRAD)/$(TRAD)-$(DATE).pat: $(TRAD)/pattern.8 $(TRAD)/pattern.rules
217 $(CAT) $(DATADIR)/$(TRAD).1 \
218 | $(SED) -e "s/@DATE@/$(DATE)/" \
219 -e "s/@GIT_VERSION@/$(GIT_VERSION)/" > $@; \
220 $(CAT) $(TRAD)/pattern.rules >> $@; \
221 $(CAT) $(DATADIR)/$(TRAD).2 >> $@; \
222 $(CAT) $(TRAD)/pattern.8 \
223 | $(ICONV) >> $@; \
224 $(CAT) $(DATADIR)/$(TRAD).3 >> $@
227 $(REFO)/pattern.8 $(REFO)/pattern.rules: $(REFO)/make-full-pattern-refo
229 $(REFO)/make-full-pattern-refo: $(REFO)/words.hyphenated.refo
230 $(CHDIR) $(REFO); \
231 $(SH) $(SCRIPTDIR)/make-full-pattern.sh $(<F) $(DATADIR)/german.tr
232 $(ECHO) done > $@
234 $(REFO)/$(REFO)-$(DATE).pat: $(REFO)/pattern.8 $(REFO)/pattern.rules
235 $(CAT) $(DATADIR)/$(REFO).1 \
236 | $(SED) -e "s/@DATE@/$(DATE)/" \
237 -e "s/@GIT_VERSION@/$(GIT_VERSION)/" > $@; \
238 $(CAT) $(REFO)/pattern.rules >> $@; \
239 $(CAT) $(DATADIR)/$(REFO).2 >> $@; \
240 $(CAT) $(REFO)/pattern.8 \
241 | $(ICONV) >> $@; \
242 $(CAT) $(DATADIR)/$(REFO).3 >> $@
245 $(SWISS)/pattern.8 $(SWISS)/pattern.rules: $(SWISS)/make-full-pattern-swiss
247 $(SWISS)/make-full-pattern-swiss: $(SWISS)/words.hyphenated.swiss
248 $(CHDIR) $(SWISS); \
249 $(SH) $(SCRIPTDIR)/make-full-pattern.sh $(<F) $(DATADIR)/german.tr
250 $(ECHO) done > $@
252 $(SWISS)/$(SWISS)-$(DATE).pat: $(SWISS)/pattern.8 $(SWISS)/pattern.rules
253 $(CAT) $(DATADIR)/$(SWISS).1 \
254 | $(SED) -e "s/@DATE@/$(DATE)/" \
255 -e "s/@GIT_VERSION@/$(GIT_VERSION)/" > $@; \
256 $(CAT) $(SWISS)/pattern.rules >> $@; \
257 $(CAT) $(DATADIR)/$(SWISS).2 >> $@; \
258 $(CAT) $(SWISS)/pattern.8 \
259 | $(ICONV) >> $@; \
260 $(CAT) $(DATADIR)/$(SWISS).3 >> $@
263 $(TRAD)/words.hyphenated.trad: $(SRCDIR)/$(WORDLIST)
264 $(CAT) $< \
265 | $(PERL) $(SCRIPTDIR)/extract-tex.pl -l -t $(PERLMAJOR) \
266 | $(SEDMAJOR) \
267 | $(SORT) > $@
269 $(REFO)/words.hyphenated.refo: $(SRCDIR)/$(WORDLIST)
270 $(CAT) $< \
271 | $(PERL) $(SCRIPTDIR)/extract-tex.pl -l $(PERLMAJOR) \
272 | $(SEDMAJOR) \
273 | $(SORT) > $@
275 $(SWISS)/words.hyphenated.swiss: $(SRCDIR)/$(WORDLIST)
276 $(CAT) $< \
277 | $(PERL) $(SCRIPTDIR)/extract-tex.pl -l -s $(PERLMAJOR) \
278 | $(SEDMAJOR) \
279 | $(SORT) > $@
282 $(TRAD)/$(TRAD)-$(DATE).tex: $(DATADIR)/$(TRAD).tex.in
283 $(CAT) $< \
284 | $(SED) -e "s/@DATE@/$(DATE)/" > $@
286 $(REFO)/$(REFO)-$(DATE).tex: $(DATADIR)/$(REFO).tex.in
287 $(CAT) $< \
288 | $(SED) -e "s/@DATE@/$(DATE)/" > $@
290 $(SWISS)/$(SWISS)-$(DATE).tex: $(DATADIR)/$(SWISS).tex.in
291 $(CAT) $< \
292 | $(SED) -e "s/@DATE@/$(DATE)/" > $@
294 pidiff:
295 $(SH) skripte/patgen-list-diff.sh $(FROM) $(TO)
297 # Listen und Patterns für Langes-S (Orthographie für Satz mit
298 # gebrochenen Schriften (de-Latf: deutsch, Latin-script-fraktur)
300 $(LANGS)/words-de-1901-Latf.txt: $(PYSCRIPTDIR)/s2long-s.py
301 $(MKDIR) $(LANGS)
302 $(PYTHON) $(PYSCRIPTDIR)/s2long-s.py -i wortliste -l de-1901 \
303 -o $(LANGS)/words-de-1901-Latf.txt
305 $(LANGS)/words-de-1901-Latf.hyphenated: $(LANGS)/words-de-1901-Latf.txt
306 $(PYTHON) $(PYSCRIPTDIR)/long_s_quasihyph.py < $< > $@
308 $(LANGS)/de-1901-Latf.pat: $(LANGS)/words-de-1901-Latf.hyphenated
309 $(CHDIR) $(LANGS); \
310 $(SH) $(SCRIPTDIR)/make-full-pattern.sh $(<F) $(DATADIR)/german.tr
311 $(CAT) $(LANGS)/pattern.8 | $(ICONV) >> $@;
313 $(LANGS)/words-de-1996-Latf.txt: $(PYSCRIPTDIR)/s2long-s.py
314 $(PYTHON) $(PYSCRIPTDIR)/s2long-s.py -i wortliste -l de-1996 \
315 -o $(LANGS)/words-de-1996-Latf.txt
317 $(LANGS)/words-de-1996-Latf.hyphenated: $(LANGS)/words-de-1996-Latf.txt
318 $(PYTHON) $(PYSCRIPTDIR)/long_s_quasihyph.py < $< > $@
321 $(LANGS)/de-1996-Latf.pat: $(LANGS)/words-de-1996-Latf.hyphenated
322 $(CHDIR) $(LANGS); \
323 $(SH) $(SCRIPTDIR)/make-full-pattern.sh $(<F) $(DATADIR)/german.tr
324 $(CAT) $(LANGS)/pattern.8 | $(ICONV) >> $@;
326 # EOF