Fine-tune parsing of "see also".

diff --git a/bots/update_sledrun_json_from_wikitext_see_also.py b/bots/update_sledrun_json_from_wikitext_see_also.py
new file mode 100644 (file)
index 0000000..4b2fdea
--- /dev/null
+++ b/bots/update_sledrun_json_from_wikitext_see_also.py
@@ -0,0 +1,143 @@
+#!/usr/bin/python
+"""
+User script for pywikibot (https://gerrit.wikimedia.org/r/pywikibot/core.git), tested with version 6.6.1.
+Put it in directory scripts/userscripts.
+
+Update a sledrun JSON page from a detail in a sledrun wikitext page.
+
+The following generators and filters are supported:
+
+&params;
+"""
+import io
+import itertools
+import json
+import re
+from itertools import takewhile, dropwhile
+from typing import Optional, List, Dict, Iterable
+
+import jsonschema
+import mwparserfromhell
+from mwparserfromhell.nodes.extras import Parameter
+
+import pywikibot
+from mwparserfromhell.nodes import Tag, Text, ExternalLink, Template, Wikilink, Heading
+from mwparserfromhell.wikicode import Wikicode
+from pywikibot import pagegenerators, Page
+from pywikibot.bot import (
+    AutomaticTWSummaryBot,
+    ConfigParserBot,
+    ExistingPageBot,
+    NoRedirectPageBot,
+    SingleSiteBot,
+)
+from pywikibot.logging import warning
+from pywikibot.site._namespace import BuiltinNamespace
+from wrpylib.json_tools import order_json_keys
+
+from wrpylib.wrmwmarkup import create_sledrun_wiki, lonlat_to_json, lonlat_ele_to_json, parse_wrmap
+from wrpylib.wrvalidators import rodelbahnbox_from_template, tristate_german_to_str, difficulty_german_to_str, \
+    avalanches_german_to_str, public_transport_german_to_str, opt_lonlat_from_str, \
+    opt_uint_from_str
+from wrpylib.lib_sledrun_wikitext_to_json import optional_set, get_sledrun_description, wikilink_to_json, \
+    external_link_to_json
+
+docuReplacements = {'&params;': pagegenerators.parameterHelp}
+
+
+class UpdateSledrunJsonFromWikiText(
+    SingleSiteBot,
+    ConfigParserBot,
+    ExistingPageBot,
+    NoRedirectPageBot,
+    AutomaticTWSummaryBot,
+):
+    def setup(self) -> None:
+        schema = Page(self.site, 'Winterrodeln:Datenschema/Rodelbahn/V1.json')
+        assert schema.content_model == 'json'
+        self.sledrun_schema = json.loads(schema.text)
+
+    def treat_page(self) -> None:
+        """Load the given page, do some changes, and save it."""
+        wikitext_content_model = 'wikitext'
+        if self.current_page.content_model != wikitext_content_model:
+            warning(f"The content model of {self.current_page.title()} is {self.current_page.content_model} "
+                    f"instead of {wikitext_content_model}.")
+            return
+
+        wikicode = mwparserfromhell.parse(self.current_page.text)
+
+        sledrun_json_page = Page(self.site, self.current_page.title() + '/Rodelbahn.json')
+        if not sledrun_json_page.exists():
+            return
+        sledrun_json = json.loads(sledrun_json_page.text)
+        sledrun_json_orig = json.loads(sledrun_json_page.text)
+        sledrun_json_orig_text = json.dumps(sledrun_json_orig, ensure_ascii=False, indent=4)
+
+        def _strip_brackets(text: str) -> str:
+            """Removes brackets if they are present"""
+            match = re.match(r'\((.+)\)', text)
+            if match is None:
+                return text
+            return match.group(1)
+
+        def _parse_weblink(line: str) -> Optional[Dict]:
+            wikicode = mwparserfromhell.parse(line)
+            nodes = dropwhile(lambda node: not isinstance(node, ExternalLink), wikicode.nodes)
+            link = next(nodes, None)
+            if link is None:
+                return
+            remaining = _strip_brackets(str(Wikicode(list(nodes))).strip()).strip()
+            title = link.title
+            if title is None:
+                if remaining != '':
+                    title = remaining
+            else:
+                title = str(link.title)
+                if remaining != '':
+                    title = f'{title} ({remaining})'
+            weblink = {'url': str(link.url)}
+            if title is not None:
+                weblink['text'] = title
+            return weblink
+
+        def _see_also(wikicode: Wikicode) -> Iterable[Dict]:
+            wikicode_common = next(iter(wikicode.get_sections(levels=[2], matches='Allgemeines')), None)
+            wikitext_common = str(wikicode_common)
+            lines = wikitext_common.split('\n')
+            lines = dropwhile(lambda line: "'''Siehe auch'''" not in line, lines)
+            lines = itertools.islice(lines, 1, None)  # omit "Siehe auch" line
+            lines = takewhile(lambda line: line.startswith('**'), lines)
+            for line in lines:
+                weblink = _parse_weblink(line)
+                if weblink is not None:
+                    yield weblink
+
+        see_also_list = list(_see_also(wikicode))
+        if len(see_also_list) > 0:
+            sledrun_json['see_also'] = see_also_list
+
+        jsonschema.validate(instance=sledrun_json, schema=self.sledrun_schema)
+        sledrun_json_ordered = order_json_keys(sledrun_json, self.sledrun_schema)
+        assert sledrun_json_ordered == sledrun_json
+        if sledrun_json == sledrun_json_orig:
+            return
+        sledrun_json_text = json.dumps(sledrun_json_ordered, ensure_ascii=False, indent=4)
+        summary = 'Information zu "Anderen Rodelbahnen" im Rodelbahn JSON aktualisiert vom Wikitext.'
+        self.userPut(sledrun_json_page, sledrun_json_orig_text, sledrun_json_text, summary=summary, contentmodel='json')
+
+
+def main(*args: str) -> None:
+    local_args = pywikibot.handle_args(args)
+    gen_factory = pagegenerators.GeneratorFactory()
+    gen_factory.handle_args(local_args)
+    gen = gen_factory.getCombinedGenerator(preload=True)
+    if gen:
+        bot = UpdateSledrunJsonFromWikiText(generator=gen)
+        bot.run()
+    else:
+        pywikibot.bot.suggest_help(missing_generator=True)
+
+
+if __name__ == '__main__':
+    main()