From 79efdb56f7b5e1e3638bda4896cdb6d9df73a690 Mon Sep 17 00:00:00 2001 From: I-Al-Istannen Date: Sun, 23 May 2021 18:22:29 +0200 Subject: [PATCH] Adjust ILIAS html explain messages --- PFERD/crawlers/ilias/kit_ilias_html.py | 6 +++--- PFERD/crawlers/ilias/kit_ilias_web_crawler.py | 6 ++++-- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/PFERD/crawlers/ilias/kit_ilias_html.py b/PFERD/crawlers/ilias/kit_ilias_html.py index cc02801..636fa68 100644 --- a/PFERD/crawlers/ilias/kit_ilias_html.py +++ b/PFERD/crawlers/ilias/kit_ilias_html.py @@ -52,12 +52,12 @@ class IliasPage: log.explain("Page is a video player, extracting URL") return self._player_to_video() if self._is_video_listing(): - log.explain("Page is a video listing, finding elements") + log.explain("Page is a video listing, searching for elements") return self._find_video_entries() if self._is_exercise_file(): - log.explain("Page is an exercise, finding elements") + log.explain("Page is an exercise, searching for elements") return self._find_exercise_entries() - log.explain("Page is a normal folder, finding elements") + log.explain("Page is a normal folder, searching for elements") return self._find_normal_entries() def _is_video_player(self) -> bool: diff --git a/PFERD/crawlers/ilias/kit_ilias_web_crawler.py b/PFERD/crawlers/ilias/kit_ilias_web_crawler.py index 1a4a529..41633d2 100644 --- a/PFERD/crawlers/ilias/kit_ilias_web_crawler.py +++ b/PFERD/crawlers/ilias/kit_ilias_web_crawler.py @@ -206,6 +206,8 @@ class KitIliasWebCrawler(HttpCrawler): raise CrawlError("Invalid course id? Didn't find anything looking like a course") # Duplicated code, but the root page is special - we want to avoid fetching it twice! + log.explain_topic("Parsing root HTML page") + log.explain(f"URL: {url}") page = IliasPage(soup, url, None) elements.extend(page.get_child_elements()) @@ -229,8 +231,8 @@ class KitIliasWebCrawler(HttpCrawler): elements.clear() async with cl: soup = await self._get_page(url) - log.explain_topic(f"Parsing HTML page for {path!r}") - log.explain(f"URL: {url!r}") + log.explain_topic(f"Parsing HTML page for {fmt_path(path)}") + log.explain(f"URL: {url}") page = IliasPage(soup, url, parent) elements.extend(page.get_child_elements())