X-Git-Url: https://pere.pagekite.me/gitweb/text-mekanikerord.git/blobdiff_plain/0927329b821f2a2e0dc09ecbd38b5269ac2d1b6d..refs/heads/master:/make-glossary diff --git a/make-glossary b/make-glossary index 32d95d6..6e36003 100755 --- a/make-glossary +++ b/make-glossary @@ -5,6 +5,8 @@ import locale from lxml import etree from lxml.etree import tostring +import json + list_topic = False filemakerxml = 'meksme-utf8.xml' @@ -55,6 +57,10 @@ for row in resultset.getchildren(): 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: return locale.strxfrm(e[lang]) @@ -69,6 +75,16 @@ def make_glossary_docbook(lang, desccodes, langcodes, output='glossary.xml'): } ) + 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: @@ -76,6 +92,7 @@ def make_glossary_docbook(lang, desccodes, langcodes, output='glossary.xml'): 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 @@ -83,18 +100,51 @@ def make_glossary_docbook(lang, desccodes, langcodes, output='glossary.xml'): e['topic'] = topicmap[lang][e['topic']] if lang in e: w = e[lang].split(',') - id = w[0] \ - .replace('[', '_') \ - .replace(']', '_') \ - .replace('(', '_') \ - .replace(')', '_') \ - .replace('/', '_') \ - .replace('\'', '_') \ - .replace(' ', '_') + id = word2id(w[0]) while id in ids: id = id + 'x' ids[id] = True - if ldesc not in e: + + # 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: @@ -120,16 +170,6 @@ def make_glossary_docbook(lang, desccodes, langcodes, output='glossary.xml'): e[codestr])))) glossary.append(entry) - # Add See also entries pointing to main entry - if 1 < len(w): - for t in w[1:]: - t = t.strip().lstrip() - #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) - def glosstermlocale(x): # Look up glossterm (FIXME figure out more robust way) t = x.getchildren()[0].text