1 files changed, 84 insertions, 0 deletions
diff --git a/bin/RPMFinder b/bin/RPMFinder
new file mode 100755
index 0000000..1797174
--- /dev/null
+++ b/bin/RPMFinder
@@ -0,0 +1,84 @@
+#!/usr/bin/env python
+
+# Searches over the network to find out which RPM package distributes
+# a given file.
+#
+# Written by Lucas C. Villa Real <lucasvr@gobolinux.org>
+# Released under the GNU GPL version 2 or above.
+
+import os
+import sys
+import argparse
+import subprocess
+from HTMLParser import HTMLParser
+
+
+class RPMFind_Parser(HTMLParser):
+    '''
+    Parses the HTML data output by rpmfind.net
+    '''
+    def __init__(self):
+	self.tags = []
+	self.candidates = []
+	HTMLParser.__init__(self)
+
+    def handle_starttag(self, tag, attrs):
+	self.tags.append(tag)
+
+    def handle_endtag(self, tag):
+	self.tags.pop()
+
+    def handle_data(self, data):
+	if len(self.tags) and self.tags[-1] == "a" and data.find(".rpm") >= 0:
+	    self.candidates.append(data)
+
+    def get_pkgname(self):
+	if len(self.candidates) == 0:
+	    return ""
+	name = os.path.commonprefix(self.candidates)
+	if name.endswith("-"):
+	    name = name[:-1]
+	return name
+
+
+class RPMFinder:
+    def find(self, path, arch, distro):
+	'''
+        Searches rpmfind.net for a given file. Arch and distro can
+	be provided to reduce the search scope. Returns the package
+        name on success or an empty string if no matches were found.
+	'''
+	self.path = path
+	self.arch = arch
+	self.distro = distro
+	return self.__search_rpmfind_net()
+
+    def __search_rpmfind_net(self):
+	path = self.path.replace("/", "%2F")
+	arch = self.arch
+	baseuri = "http://rpmfind.net/linux/rpm2html/search.php"
+	query = "?query={0}&submit=Search+...&system=&arch={1}".format(path, arch)
+	html = subprocess.check_output(["wget", "--quiet", "{0}{1}".format(baseuri, query), "-O", "-"])
+
+	htmlparser = RPMFind_Parser()
+        htmlparser.feed(html)
+	return htmlparser.get_pkgname()
+
+
+def main():
+    argparser = argparse.ArgumentParser(argument_default="")
+    argparser.add_argument("--path", type=str, help="File name to search for in the remote RPM databases")
+    argparser.add_argument("--arch", type=str, help="Architecture (optional)")
+    argparser.add_argument("--distro", type=str, help="Distribution (optional)")
+    args = argparser.parse_args()
+
+    if len(args.path) == 0:
+	argparser.print_help()
+	sys.exit(1)
+
+    pkgname = RPMFinder().find(args.path, args.arch, args.distro)
+    if len(pkgname):
+	print pkgname
+
+if __name__ == "__main__":
+    main()