Korrigert nordsamisk tekst basert på tilbakemelding fra Børre Gaup.

[text-mekanikerord.git] / make-glossary

diff --git a/make-glossary b/make-glossary
index 6095a008377e50145794ad51c9f3d06b70486a27..6e36003a21f81ba14ded2f503322c535c7ce6353 100755 (executable)
--- a/make-glossary
+++ b/make-glossary
@@ -1,19 +1,47 @@
 #!/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
 

-tree = etree.parse('mekanikk-1999/meksme-utf8.xml')
+import json
+
+list_topic = False
+
+filemakerxml = 'meksme-utf8.xml'
+
+tree = etree.parse(filemakerxml)

 root = tree.getroot()
 
 #print(root)
 #print(tostring(tree))
 
 cols = (
 root = tree.getroot()
 
 #print(root)
 #print(tostring(tree))
 
 cols = (

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

 )
 
 )
 

+topicmap = {
+    'nb' : {
+        'fáddá': 'tema',
+        'ávnnas': 'emne',
+        'eanan': 'land',
+        'biras': 'miljø',
+        'huksen': 'bygg',
+        'bohcci': 'rør',
+        'data': 'data',
+        'hydr': 'hydraulikk',
+        'fys': 'fysikk',
+        'sveis': 'sveising',
+        'mihttu': 'måling',
+        'elektro': 'elektro',
+        'neavvu': 'verktøy',
+        'mohtor': 'motor',
+        'mašiidna': 'maskin',
+        'fuolahas': 'bearbeiding',
+    }
+}
+

 resultset = root.find("{http://www.filemaker.com/fmpxmlresult}RESULTSET")
 
 words = []
 resultset = root.find("{http://www.filemaker.com/fmpxmlresult}RESULTSET")
 
 words = []


@@ -30,37 +58,164 @@ for row in resultset.getchildren():
     #print(d)
     words.append(d)
 
     #print(d)
     words.append(d)
 

-def make_glossary(lang):
-    print(".. glossary::")
-    print()
+    with open('meksme-utf8.json', 'w') as f:
+        json.dump(words, f)

 
 

-    def langsort(e):
+def langsort(lang, e):
+    if lang in e:
+        return locale.strxfrm(e[lang])
+    else:
+        return locale.strxfrm(e['se'])
+
+def make_glossary_docbook(lang, desccodes, langcodes, output='glossary.xml'):
+    import lxml.builder
+    E = lxml.builder.ElementMaker(
+        nsmap={
+#            'xi': "http://www.w3.org/2001/XInclude",
+        }
+    )
+
+    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)))
+    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:
         if lang in e:

-            return e[lang]
-        else:
-            return e['sme']
-    for e in sorted(words, key=langsort):
-        if lang in e and 'desc-%s' % lang in e:
-            if 'topic' not in e:
-                e['topic'] = 'n/a'
-            #print(e)
-            print("  %s [%s]\n     %s" % (e[lang], e['topic'], e['desc-%s' % lang]))
-            print()
+            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]))
+            indexit(entry, e[lang])
+            lstr = ""
+            for l in langcodes:
+                if l != lang and l in e:
+                    lstr += "%s (%s) " % (e[l], l)
+                    # Add foreign words to index, split on comma
+                    indexit(entry, e[l], l)
+            if "" != lstr:
+                entry.append(E.glossdef(E.para(lstr)))
+            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)
+
+    def glosstermlocale(x):
+        # Look up glossterm (FIXME figure out more robust way)
+        t = x.getchildren()[0].text
+        if t:
+            return locale.strxfrm(t)

         else:
         else:

-            # ERROR / missing definition
-            pass
-
-print("Nordsamisk")
-print("==========")
-print()
-make_glossary(lang='sme')
-
-print("Norsk")
-print("=====")
-print()
-make_glossary(lang='nb')
-
-#print("Engelsk")
-#print("=====")
-#print()
-#make_glossary(lang='en')
+            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)
+    with open(output, 'wb') as f:
+        f.write(content)
+
+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' == args.langcode:
+    print("Norsk/bokmål")
+    print()
+    make_glossary_docbook(lang='nb', desccodes=('nb',),
+                          langcodes=('en', 'se', 'sv', 'da', 'fi', 'is',),
+                          output=args.output)
+elif 'se' == args.langcode:
+    print("Nordsamisk")
+    print()
+    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()
+    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)