[gitiles] Add parse_repo_url function

Add a function to parse a Gitiles repo URL supporting various edge cases.

R=tandrii@chromium.org

Bug: 877161
Change-Id: Ib4d6a28a2ace2e594f1a4ca31460a446fc6fa052
Reviewed-on: https://chromium-review.googlesource.com/1204914
Commit-Queue: Nodir Turakulov <nodir@chromium.org>
Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>

recipes/recipe_modules/gitiles/api.py

@@ -3,6 +3,7 @@
 # found in the LICENSE file.
 
 import base64
+import urlparse
 
 from recipe_engine import recipe_api
 
@@ -200,3 +201,40 @@ def download_archive(self, repository_url, destination,
       ex = self.m.step.StepFailure(step_name)
       ex.gitiles_skipped_files = stat['names']
       raise ex
+
+  def parse_repo_url(self, repo_url):
+    """Returns (host, project) pair.
+
+    Returns (None, None) if repo_url is not recognized.
+    """
+    return parse_repo_url(repo_url)
+
+
+def parse_http_host_and_path(url):
+  # Copied from https://chromium.googlesource.com/infra/luci/recipes-py/+/809e57935211b3fcb802f74a7844d4f36eff6b87/recipe_modules/buildbucket/util.py
+  parsed = urlparse.urlparse(url)
+  if not parsed.scheme:
+    parsed = urlparse.urlparse('https://' + url)
+  if (parsed.scheme in ('http', 'https') and
+      not parsed.params and
+      not parsed.query and
+      not parsed.fragment):
+    return parsed.netloc, parsed.path
+  return None, None
+
+
+def parse_repo_url(repo_url):
+  """Returns (host, project) pair.
+
+  Returns (None, None) if repo_url is not recognized.
+  """
+  # Adapted from https://chromium.googlesource.com/infra/luci/recipes-py/+/809e57935211b3fcb802f74a7844d4f36eff6b87/recipe_modules/buildbucket/util.py
+  host, project = parse_http_host_and_path(repo_url)
+  if not host or not project or '+' in project.split('/'):
+    return None, None
+  project = project.strip('/')
+  if project.startswith('a/'):
+    project = project[len('a/'):]
+  if project.endswith('.git'):
+    project = project[:-len('.git')]
+  return host, project

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/a prefix.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "host",
+      "path/to/project"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/basic.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "host",
+      "path/to/project"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/git suffix.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "host",
+      "path/to/project"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/http and a prefix.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "host",
+      "path/to/project"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/http.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "host",
+      "path/to/project"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/no scheme.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "host",
+      "path/to/project"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/plus.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "None",
+      "None"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.expected/query string param.json

@@ -0,0 +1,15 @@
+[
+  {
+    "cmd": [
+      "echo",
+      "None",
+      "None"
+    ],
+    "name": "build"
+  },
+  {
+    "name": "$result",
+    "recipe_result": null,
+    "status_code": 0
+  }
+]

recipes/recipe_modules/gitiles/tests/parse_repo_url.py

@@ -0,0 +1,30 @@
+# Copyright 2018 The LUCI Authors. All rights reserved.
+# Use of this source code is governed under the Apache License, Version 2.0
+# that can be found in the LICENSE file.
+
+DEPS = [
+  'gitiles',
+  'recipe_engine/properties',
+  'recipe_engine/step',
+]
+
+
+def RunSteps(api):
+  repo_url = api.properties['repo_url']
+  host, project = api.gitiles.parse_repo_url(repo_url)
+  api.step('build', ['echo', str(host), str(project)])
+
+
+def GenTests(api):
+
+  def case(name, repo_url):
+    return api.test(name) + api.properties(repo_url=repo_url)
+
+  yield case('basic', 'https://host/path/to/project')
+  yield case('http', 'http://host/path/to/project')
+  yield case('a prefix', 'https://host/a/path/to/project')
+  yield case('git suffix', 'https://host/path/to/project.git')
+  yield case('http and a prefix', 'http://host/a/path/to/project')
+  yield case('no scheme', 'host/a/path/to/project')
+  yield case('query string param', 'https://host/a/path/to/project?a=b')
+  yield case('plus', 'https://host/path/to/project/+/master')