refactor: split web exporter for offline testing

Extract prepare/publish/build_compilation and dumpall_web so offline tests
can assert web output without writing pages alone.

Signed-off-by: Alan Peixinho <alan.peixinho@profusion.mobi>

Author: alanpeixinho

regzbot/export_web.py

@@ -17,7 +17,6 @@
 logger = regzbot.logger
 
 
-
 def _web_now():
     mail_now = regzbot._TESTING.get("mail_now")
     if mail_now is not None:
@@ -799,7 +798,7 @@ def create_individual_page(cls, htmlpages, unhandled_count, regression):
             )
 
     @classmethod
-    def createpage_compilation(cls, htmlpages, unhandled_count, categories, pagename):
+    def build_compilation(cls, htmlpages, unhandled_count, categories, pagename):
         tablecolumns = 3
         yattagdoc = yattag.Doc()
         cls.outpage_head(yattagdoc)
@@ -830,8 +829,12 @@ def createpage_compilation(cls, htmlpages, unhandled_count, categories, pagename
                                 with yattagdoc.tag("td", style="width: 100px;"):
                                     yattagdoc.text(regressionweb.treename)
             cls.outpage_footer(yattagdoc, unhandled_count)
+        return yattagdoc
 
-            cls.outpage_write(pagename, yattagdoc)
+    @classmethod
+    def createpage_compilation(cls, htmlpages, unhandled_count, categories, pagename):
+        yattagdoc = cls.build_compilation(htmlpages, unhandled_count, categories, pagename)
+        cls.outpage_write(pagename, yattagdoc)
 
     @classmethod
     def create_events(cls, directory, unhandled_count, htmlpages, eventslist):
@@ -1109,24 +1112,22 @@ def regression_to_json(cls, regression):
         }
 
     @classmethod
-    def compile(cls):
-        logger.debug("[webpages] generating")
+    def _events_gmtime_offset(cls):
+        events_gmtime_offset = int(_web_now().timestamp()) - 604800
+        if regzbot.is_running_citesting("offline"):
+            events_gmtime_offset = int(_web_now().timestamp()) - 604800 * 52 * 10
+        return events_gmtime_offset
 
+    @classmethod
+    def prepare(cls):
         # these are the pages we are going to create
         htmlpages = ("next", "mainline", "stable", "new", "all", "resolved", "inconclusive")
-
-        # handle this page first, as we need something from it anyway
-        unhandled_count = cls.create_unhandled(regzbot.WEBPAGEDIR, htmlpages)
+        unhandled_count = len(list(UnhandledEventWeb.get_all()))
+        events_gmtime_offset = cls._events_gmtime_offset()
 
         # gather everything we need
         regressionslist = list()
         eventslist = list()
-        events_gmtime_offset = (
-            int(_web_now().timestamp()) - 604800
-        )
-        if regzbot.is_running_citesting("offline"):
-            events_gmtime_offset = 604800 * 52 * 10
-
         json_data = list()
         for regression in RegressionWeb.get_all():
             json_data.append(cls.regression_to_json(regression))
@@ -1164,12 +1165,28 @@ def compile(cls):
                 )
             )
 
-        cls.create_scriptfile_reldate()
-
         eventslist.sort(key=lambda x: x["gmtime"], reverse=True)
+        return {
+            "htmlpages": htmlpages,
+            "unhandled_count": unhandled_count,
+            "regressionslist": regressionslist,
+            "json_data": json_data,
+            "eventslist": eventslist,
+        }
+
+    @classmethod
+    def publish(cls, prepared):
+        htmlpages = prepared["htmlpages"]
+        unhandled_count = prepared["unhandled_count"]
+        regressionslist = prepared["regressionslist"]
+        json_data = prepared["json_data"]
+        eventslist = prepared["eventslist"]
+
+        # handle this page first, as we need something from it anyway
+        cls.create_unhandled(regzbot.WEBPAGEDIR, htmlpages)
+        cls.create_scriptfile_reldate()
         cls.create_events(regzbot.WEBPAGEDIR, unhandled_count, htmlpages, eventslist)
         # we don't need this anymore now that we iterated over all regressions
-        eventslist = events_gmtime_offset = None
 
         # create the page listing all regressions, sorted by date
         regressionslist.sort(key=lambda x: x.gmtime_report, reverse=True)
@@ -1230,4 +1247,24 @@ def compile(cls):
         with open(os.path.join(regzbot.WEBPAGEDIR, "regressions.json"), "w") as jsonfile:
             jsonfile.write(json.dumps(json_data))
 
+    @classmethod
+    def compile(cls):
+        logger.debug("[webpages] generating")
+        cls.publish(cls.prepare())
         logger.debug("[webpages] generated")
+
+
+def dumpall_web():
+    prepared = RegExportWeb.prepare()
+    regressionslist = prepared["regressionslist"]
+    regressionslist.sort(key=lambda x: x.gmtime_activity, reverse=True)
+    categories = RegExportWeb.categorize(regressionslist)
+    mainline = categories.get("mainline", {})
+    if not any(section["entries"] for section in mainline.values()):
+        return
+
+    doc = RegExportWeb.build_compilation(
+        prepared["htmlpages"], prepared["unhandled_count"], mainline, "mainline"
+    )
+    yield yattag.indent(doc.getvalue())
+    yield "\n"