From c8a075fc7a49eef91a026e59dbfdf23fb4e612e8 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Fran=C3=A7ois=20Poulain?= <fpoulain@metrodore.fr>
Date: Sat, 8 Aug 2020 13:45:45 +0200
Subject: [PATCH] feat(convert): implante les filtres html autobr et autop de
 drupal

---
 drupal2spip_lal/base/convert.py               | 55 +++++++++++++++++--
 .../base/management/commands/import.py        |  1 -
 2 files changed, 49 insertions(+), 7 deletions(-)

diff --git a/drupal2spip_lal/base/convert.py b/drupal2spip_lal/base/convert.py
index e00af0d..4291b82 100644
--- a/drupal2spip_lal/base/convert.py
+++ b/drupal2spip_lal/base/convert.py
@@ -110,10 +110,48 @@ def fetch_and_remove_logo(article, force_download):
         article.save()
 
 
-def sanitarize_html(html):
+def filter_html(html):
+    def auto_p(html):
+        re_paragraph = re.compile(r'\s*\n\s*\n\s*')
+        soup = bs(html, 'html.parser')
+        for string in soup.find_all(string=re_paragraph):
+            new_item = bs('', 'html.parser')
+            for substring in [s for s in re_paragraph.split(string) if s]:
+                p = new_item.new_tag('p')
+                p.string = substring.strip()
+                new_item.append(p)
+            string.replace_with(new_item)
+        return str(soup)
+
+    def auto_br(html):
+        re_break = re.compile(r'\s*\n\s*')
+        soup = bs(html, 'html.parser')
+        for string in soup.find_all(string=re_break):
+            new_item = bs('', 'html.parser')
+            for index, substring in enumerate([s for s in re_break.split(string) if s]):
+                if index:
+                    br = new_item.new_tag('br')
+                    new_item.append(br)
+                new_item.append(substring.strip())
+            string.replace_with(new_item)
+        return str(soup)
+
+    html = auto_p(html)
+    html = auto_br(html)
+    return html
+
+
+def sanitarize_html(html, node_fmt):
     html = strong_to_dl(html)
-    soup = bs(html, 'html.parser')
-    return str(soup)
+
+    if node_fmt == 'PHP code':
+        raise NotImplementedError("Ce node est au format PHP.")
+    if node_fmt == 'Filtered HTML':
+        html = filter_html(html)
+    if node_fmt == 'Filtered HTML':
+        html = str(bs(html, 'html.parser'))
+
+    return html
 
 
 def convert_node(node, options):
@@ -139,7 +177,8 @@ def convert_node(node, options):
             or len(set(spip_urls.values_list('id_objet', flat=True))) != 1
             or spip.Urls.objects.filter(
                 type='article', id_objet=article_id
-            ).count() != spip_urls.count()
+            ).count()
+            != spip_urls.count()
         ):
             # incohérence dans les urls
             raise ValueError(
@@ -147,15 +186,19 @@ def convert_node(node, options):
                 "de s'assurer qu'il s'agisse d'un import prééexistant."
             )
 
+    node_fmt = node.published_revision.format.name
+
     article = None
     article_attributes = {
         'date': convert_timestamp(node.published_revision.timestamp),
         'date_modif': convert_timestamp(node.changed),
         'date_redac': convert_timestamp(node.created),
-        'descriptif': sanitarize_html(node.published_revision.teaser),
+        'descriptif': sanitarize_html(
+            node.published_revision.teaser, node_fmt
+        ),
         'maj': convert_timestamp(node.changed),
         'statut': 'publie' if node.status else 'prepa',
-        'texte': sanitarize_html(node.published_revision.body),
+        'texte': sanitarize_html(node.published_revision.body, node_fmt),
         'titre': node.title,
     }
     if not spip_urls.exists():
diff --git a/drupal2spip_lal/base/management/commands/import.py b/drupal2spip_lal/base/management/commands/import.py
index dca7c59..5136b91 100644
--- a/drupal2spip_lal/base/management/commands/import.py
+++ b/drupal2spip_lal/base/management/commands/import.py
@@ -5,7 +5,6 @@ from django.core.management.base import BaseCommand
 from drupal2spip_lal.base.convert import convert_node
 from drupal2spip_lal.drupal.models import Node
 
-
 logger = logging.getLogger('drupal2spip_lal')