depgraph: split out similar_name_search func

This will allow the code to be re-used for bug #444596. Copyright
begins in 2011 since that's when the code for this feature was first
added in commit aa78cc8da18015b7d1e4eec277b5a7f940fe357c.

1 files changed, 57 insertions, 0 deletions

diff --git a/pym/portage/dbapi/_similar_name_search.py b/pym/portage/dbapi/_similar_name_search.py
new file mode 100644
index 000000000..d569fbf89
--- /dev/null
+++ b/pym/portage/dbapi/_similar_name_search.py
@@ -0,0 +1,57 @@
+# Copyright 2011-2012 Gentoo Foundation
+# Distributed under the terms of the GNU General Public License v2
+
+import difflib
+
+from portage.versions import catsplit
+
+def similar_name_search(dbs, atom):
+
+	cp = atom.cp.lower()
+	cat, pkg = catsplit(cp)
+	if cat == "null":
+		cat = None
+
+	all_cp = set()
+	for db in dbs:
+		all_cp.update(db.cp_all())
+
+	# discard dir containing no ebuilds
+	all_cp.discard(cp)
+
+	orig_cp_map = {}
+	for cp_orig in all_cp:
+		orig_cp_map.setdefault(cp_orig.lower(), []).append(cp_orig)
+	all_cp = set(orig_cp_map)
+
+	if cat:
+		matches = difflib.get_close_matches(cp, all_cp)
+	else:
+		pkg_to_cp = {}
+		for other_cp in list(all_cp):
+			other_pkg = catsplit(other_cp)[1]
+			if other_pkg == pkg:
+				# Check for non-identical package that
+				# differs only by upper/lower case.
+				identical = True
+				for cp_orig in orig_cp_map[other_cp]:
+					if catsplit(cp_orig)[1] != \
+						catsplit(atom.cp)[1]:
+						identical = False
+						break
+				if identical:
+					# discard dir containing no ebuilds
+					all_cp.discard(other_cp)
+					continue
+			pkg_to_cp.setdefault(other_pkg, set()).add(other_cp)
+
+		pkg_matches = difflib.get_close_matches(pkg, pkg_to_cp)
+		matches = []
+		for pkg_match in pkg_matches:
+			matches.extend(pkg_to_cp[pkg_match])
+
+	matches_orig_case = []
+	for cp in matches:
+		matches_orig_case.extend(orig_cp_map[cp])
+
+	return matches_orig_case