Korrigert nordsamisk tekst basert på tilbakemelding fra Børre Gaup.

[text-mekanikerord.git] / make-glossary

diff --git a/make-glossary b/make-glossary
index c54b02c694dcfaaaef5830b27b680084898b7a3a..6e36003a21f81ba14ded2f503322c535c7ce6353 100755 (executable)
--- a/make-glossary
+++ b/make-glossary
@@ -1,8 +1,14 @@
 #!/usr/bin/python3
 
 #!/usr/bin/python3
 

+import locale
+

 from lxml import etree
 from lxml.etree import tostring
 
 from lxml import etree
 from lxml.etree import tostring
 

+import json
+
+list_topic = False
+

 filemakerxml = 'meksme-utf8.xml'
 
 tree = etree.parse(filemakerxml)
 filemakerxml = 'meksme-utf8.xml'
 
 tree = etree.parse(filemakerxml)


@@ -12,7 +18,7 @@ root = tree.getroot()
 #print(tostring(tree))
 
 cols = (
 #print(tostring(tree))
 
 cols = (

-    'topic', 'sme', 'desc-sme', 'desc-nb', 'nb', 'sv', 'fi', 'en', 'is',
+    'topic', 'se', 'desc-se', 'desc-nb', 'nb', 'sv', 'fi', 'en', 'is',

 )
 
 topicmap = {
 )
 
 topicmap = {


@@ -51,13 +57,17 @@ for row in resultset.getchildren():
         index += 1
     #print(d)
     words.append(d)
         index += 1
     #print(d)
     words.append(d)

+
+    with open('meksme-utf8.json', 'w') as f:
+        json.dump(words, f)
+

 def langsort(lang, e):
     if lang in e:
 def langsort(lang, e):
     if lang in e:

-        return e[lang]
+        return locale.strxfrm(e[lang])

     else:
     else:

-        return e['sme']
+        return locale.strxfrm(e['se'])

 
 

-def make_glossary_docbook(lang, langcodes):
+def make_glossary_docbook(lang, desccodes, langcodes, output='glossary.xml'):

     import lxml.builder
     E = lxml.builder.ElementMaker(
         nsmap={
     import lxml.builder
     E = lxml.builder.ElementMaker(
         nsmap={


@@ -65,20 +75,79 @@ def make_glossary_docbook(lang, langcodes):
         }
     )
 
         }
     )
 

+    def word2id(word):
+        return word \
+                .replace('[', '_') \
+                .replace(']', '_') \
+                .replace('(', '_') \
+                .replace(')', '_') \
+                .replace('/', '_') \
+                .replace('\'', '_') \
+                .replace(' ', '_')
+

     def indexit(entry, wlist, lang=None):
         for w in wlist.split(","):
             if "" != w:
                 if lang and '[' not in w:
                     w += "[%s]" % lang
                 entry.append(E.indexterm(E.primary(w)))
     def indexit(entry, wlist, lang=None):
         for w in wlist.split(","):
             if "" != w:
                 if lang and '[' not in w:
                     w += "[%s]" % lang
                 entry.append(E.indexterm(E.primary(w)))

-    glossary = E.glosslist()
+    ids = {}
+    redirects = {}
+    glossary = E.glossary()

     for e in sorted(words, key=lambda x: langsort(lang, x)):
         ldesc = 'desc-%s' % lang
         if 'topic' in e and lang in topicmap:
             e['topic'] = topicmap[lang][e['topic']]
         if lang in e:
     for e in sorted(words, key=lambda x: langsort(lang, x)):
         ldesc = 'desc-%s' % lang
         if 'topic' in e and lang in topicmap:
             e['topic'] = topicmap[lang][e['topic']]
         if lang in e:

-            entry = E.glossentry()
-            if 'topic' in e:
+            w = e[lang].split(',')
+            id = word2id(w[0])
+            while id in ids:
+                id = id + 'x'
+            ids[id] = True
+
+            # First handle redirections with not extra info
+            if -1 != e[lang].find('>') and ldesc not in e:
+                p = e[lang].split(' > ')
+                if p[0] in redirects: # Skip if already added
+                    continue
+                if -1 == p[1].find(','):
+                    if '-' == p[1][-1]:
+                        print("warning: Skipping dangling reference %s -> %s" %
+                              (p[0], p[1]))
+                    else:
+                        seeentry = E.glossentry()
+                        seeentry.append(E.glossterm(p[0]))
+                        id = word2id(p[1])
+                        seeentry.append(E.glosssee(otherterm=id))
+                        glossary.append(seeentry)
+                        redirects[p[0]] = id
+                else:
+                  print("warning: skipping split refererence %s -> %s" %
+                        (p[0], p[1]))
+                  if False: # Not allowed in docbook
+                    seeentry = E.glossentry()
+                    seeentry.append(E.glossterm(p[0]))
+                    for s in p[1].split(','):
+                        s = s.strip().lstrip()
+                        seeentry.append(E.glosssee(otherterm=word2id(s)))
+                    glossary.append(seeentry)
+                continue
+
+            # Add See also entries pointing to main entry
+            if 1 < len(w):
+                for t in w[1:]:
+                    t = t.strip().lstrip()
+                    if t not in redirects:
+                    #print("info: Adding see also entry for %s" % t)
+                        seeentry = E.glossentry()
+                        seeentry.append(E.glossterm(t))
+                        seeentry.append(E.glosssee(otherterm=id))
+                        glossary.append(seeentry)
+                        redirects[t] = id
+            elif ldesc not in e:
+                print("warning: term %s missing primary language %s description" % (e[lang], lang))
+            entry = E.glossentry(id=id)
+            if list_topic and 'topic' in e:

                 entry.append(E.glossterm('%s [%s]' % (e[lang], e['topic'])))
             else:
                 entry.append(E.glossterm(e[lang]))
                 entry.append(E.glossterm('%s [%s]' % (e[lang], e['topic'])))
             else:
                 entry.append(E.glossterm(e[lang]))


@@ -91,38 +160,62 @@ def make_glossary_docbook(lang, langcodes):
                     indexit(entry, e[l], l)
             if "" != lstr:
                 entry.append(E.glossdef(E.para(lstr)))
                     indexit(entry, e[l], l)
             if "" != lstr:
                 entry.append(E.glossdef(E.para(lstr)))

-            if ldesc in e:
-                entry.append(E.glossdef(E.para(e[ldesc])))
+            else:
+                # only single word witout translations, skip it
+                continue
+            for desccode in desccodes:
+                codestr = 'desc-%s' % desccode
+                if codestr in e:
+                    entry.append(E.glossdef(E.para("(%s): %s" % (desccode,
+                                                               e[codestr]))))

             glossary.append(entry)
 
             glossary.append(entry)
 

-    if False: # failed to set docbook glossary like xmlto and lint want it...
-      glossary =\
-        E.glossary(E.title("x"),
-                   E.glossdiv(E.title("y"),
-                              glossary))
+    def glosstermlocale(x):
+        # Look up glossterm (FIXME figure out more robust way)
+        t = x.getchildren()[0].text
+        if t:
+            return locale.strxfrm(t)
+        else:
+            return ""
+    # Sort list to mix seealso entries into their correct location.
+    glossary[:] = sorted(glossary, key=glosstermlocale)

 
 

+    l = len(glossary)
+    print("info: dictionary contain %d entries" % l)
+    

     content = lxml.etree.tostring(glossary,
                                   pretty_print=True,
                                   xml_declaration=True,
                                   encoding='UTF-8')
 #    print(content)
     content = lxml.etree.tostring(glossary,
                                   pretty_print=True,
                                   xml_declaration=True,
                                   encoding='UTF-8')
 #    print(content)

-    with open('glossary.xml', 'wb') as f:
+    with open(output, 'wb') as f:

         f.write(content)
 
         f.write(content)
 

-focus = 'nb'
-#focus = 'sme'
-#focus = 'sv'
-#focus = 'en'
+import argparse
+parser = argparse.ArgumentParser()
+parser.add_argument("langcode", help="language code to generate glossary for")
+parser.add_argument("--output", help="where to store the glossary")
+args = parser.parse_args()
+
+locale.setlocale(locale.LC_ALL, '')

 
 

-if 'nb' == focus:
+if 'nb' == args.langcode:

     print("Norsk/bokmål")
     print()
     print("Norsk/bokmål")
     print()

-    make_glossary_docbook(lang='nb', langcodes=('en', 'sme', 'sv', 'da', 'fi', 'is',))
-elif 'sme' == focus:
+    make_glossary_docbook(lang='nb', desccodes=('nb',),
+                          langcodes=('en', 'se', 'sv', 'da', 'fi', 'is',),
+                          output=args.output)
+elif 'se' == args.langcode:

     print("Nordsamisk")
     print()
     print("Nordsamisk")
     print()

-    make_glossary_docbook(lang='sme', langcodes=('nb', 'en', 'sv', 'da', 'fi', 'is',))
-elif 'en' == focus:
+    make_glossary_docbook(lang='se', desccodes=('se', 'nb'),
+                          langcodes=('nb', 'en', 'sv', 'da', 'fi', 'is',),
+                          output=args.output)
+elif 'en' == args.langcode:

     print("Engelsk")
     print()
     print("Engelsk")
     print()

-    make_glossary_docbook(lang='en',  langcodes=('en', 'nb', 'sme', 'sv', 'da', 'fi', 'is',))
+    make_glossary_docbook(lang='en', desccodes=('en'),
+                          langcodes=('en', 'nb', 'se', 'sv', 'da', 'fi', 'is',),
+                          output=args.output)
+else:
+    print("error: Unknown language code %s" % args.langcode)