euscan: add kde handler and fix some regex issues

Signed-off-by: Corentin Chary <corentincj@iksaif.net>

pym/euscan/handlers/__init__.py

@@ -1,9 +1,6 @@
-from euscan.handlers import generic
+from euscan.handlers import generic, php, pypi, rubygem, kde
-from euscan.handlers import php
-from euscan.handlers import pypi
-from euscan.handlers import rubygem
 
-handlers = [ php, pypi, rubygem, generic ]
+handlers = [ kde, php, pypi, rubygem, generic ]
 
 def find_best_handler(cpv, url):
     for handler in handlers:

pym/euscan/handlers/generic.py

@@ -18,6 +18,7 @@ def scan_html(data, url, pattern):
         href = link.get("href")
         if not href:
             continue
+
         if href.startswith(url):
             href = href.replace(url, "", 1)
 

pym/euscan/handlers/kde.py

@@ -0,0 +1,34 @@
+from euscan.handlers import generic
+
+def can_handle(cpv, url):
+    if url.startswith('mirror://kde/'):
+        return True
+    return False
+
+def clean_results(results):
+    ret = []
+
+    for path, version in results:
+        if version == '5SUMS':
+            continue
+        ret.append((path, version))
+
+    return ret
+
+def scan(cpv, url):
+    results = generic.scan(cpv, url)
+
+    if url.startswith('mirror://kde/unstable/'):
+        url = url.replace('mirror://kde/unstable/', 'mirror://kde/stable/')
+        results += generic.scan(cpv, url)
+
+    return clean_results(results)
+
+def brute_force(cpv, url):
+    results = generic.brute_force(cpv, url)
+
+    if url.startswith('mirror://kde/unstable/'):
+        url = url.replace('mirror://kde/unstable/', 'mirror://kde/stable/')
+        results += generic.brute_force(cpv, url)
+
+    return clean_results(results)

pym/euscan/helpers.py

@@ -334,16 +334,26 @@ def tryurl(fileurl, template):
     return result
 
 def regex_from_template(template):
+    # Escape
     template = re.escape(template)
+
+    # Unescape specific stuff
     template = template.replace('\$\{', '${')
     template = template.replace('\}', '}')
     template = template.replace('}\.$', '}.$')
-    template = template.replace('${1}', r'([\d]+?)')
+
-    template = re.sub(r'(\$\{\d+\}\.?)+', r'([\w]+?)', template)
+    # Replace ${\d+}
+    #template = template.replace('${0}', r'([\d]+?)')
+    template = re.sub(r'(\$\{\d+\}(\.?))+', r'([\w\.]+?)', template)
+
     #template = re.sub(r'(\$\{\d+\}\.?)+', r'([\w]+?)', template)
     #template = re.sub(r'(\$\{\d+\}\.+)+', '(.+?)\.', template)
     #template = re.sub(r'(\$\{\d+\})+', '(.+?)', template)
+
+    # Full version
     template = template.replace('${PV}', _v)
+
+    # End
     template = template + r'/?$'
     return template
 
@@ -372,6 +382,7 @@ def generate_scan_paths(url):
         else:
             path += "/"
             path += chunk
+
     return steps
 
 def parse_mirror(uri):