implement parsing human-readable names from URLs

Based off of commit 257b768

Upstream-PR: NixOS/nix#8678
Co-authored-by: Felix Uhl <[email protected]>
Change-Id: Idcb7f6191ca3310ef9dc854197f7798260c3f71d

src/libutil/meson.build

@@ -31,6 +31,7 @@ libutil_sources = files(
   'tarfile.cc',
   'thread-pool.cc',
   'url.cc',
+  'url-name.cc',
   'util.cc',
   'xml-writer.cc',
 )
@@ -92,6 +93,7 @@ libutil_headers = files(
   'topo-sort.hh',
   'types.hh',
   'url-parts.hh',
+  'url-name.hh',
   'url.hh',
   'util.hh',
   'variant-wrapper.hh',

src/libutil/url-name.cc

@@ -0,0 +1,59 @@
+#include <iostream>
+#include <regex>
+
+#include "url-name.hh"
+
+namespace nix {
+
+static std::string const attributeNamePattern("[a-z0-9_-]+");
+static std::regex const lastAttributeRegex("(?:" + attributeNamePattern + "\\.)*(?!default)(" + attributeNamePattern +")(\\^.*)?");
+static std::string const pathSegmentPattern("[a-zA-Z0-9_-]+");
+static std::regex const lastPathSegmentRegex(".*/(" + pathSegmentPattern +")");
+static std::regex const secondPathSegmentRegex("(?:" + pathSegmentPattern + ")/(" + pathSegmentPattern +")(?:/.*)?");
+static std::regex const gitProviderRegex("github|gitlab|sourcehut");
+static std::regex const gitSchemeRegex("git($|\\+.*)");
+static std::regex const defaultOutputRegex(".*\\.default($|\\^.*)");
+
+std::optional<std::string> getNameFromURL(ParsedURL const & url)
+{
+    std::smatch match;
+
+    /* If there is a dir= argument, use its value */
+    if (url.query.count("dir") > 0) {
+        return url.query.at("dir");
+    }
+
+    /* If the fragment isn't a "default" and contains two attribute elements, use the last one */
+    if (std::regex_match(url.fragment, match, lastAttributeRegex)) {
+        return match.str(1);
+    }
+
+    /* If this is a github/gitlab/sourcehut flake, use the repo name */
+    if (
+        std::regex_match(url.scheme, gitProviderRegex)
+        && std::regex_match(url.path, match, secondPathSegmentRegex)
+    ) {
+        return match.str(1);
+    }
+
+    /* If it is a regular git flake, use the directory name */
+    if (
+        std::regex_match(url.scheme, gitSchemeRegex)
+        && std::regex_match(url.path, match, lastPathSegmentRegex)
+    ) {
+        return match.str(1);
+    }
+
+    /* If everything failed but there is a non-default fragment, use it in full */
+    if (!url.fragment.empty() && !std::regex_match(url.fragment, defaultOutputRegex))
+        return url.fragment;
+
+    /* If there is no fragment, take the last element of the path */
+    if (std::regex_match(url.path, match, lastPathSegmentRegex))
+        return match.str(1);
+
+    /* If even that didn't work, the URL does not contain enough info to determine a useful name */
+    return {};
+}
+
+}

src/libutil/url-name.hh

@@ -0,0 +1,26 @@
+#pragma once
+///@file url-name.hh, for some hueristic-ish URL parsing.
+
+#include <string>
+#include <optional>
+
+#include "url.hh"
+#include "url-parts.hh"
+#include "util.hh"
+#include "split.hh"
+
+namespace nix {
+
+/**
+ * Try to extract a reasonably unique and meaningful, human-readable
+ * name of a flake output from a parsed URL.
+ * When nullopt is returned, the callsite should use information available
+ * to it outside of the URL to determine a useful name.
+ * This is a heuristic approach intended for user interfaces.
+ * @return nullopt if the extracted name is not useful to identify a
+ * flake output, for example because it is empty or "default".
+ * Otherwise returns the extracted name.
+ */
+std::optional<std::string> getNameFromURL(ParsedURL const & url);
+
+}

src/libutil/url-parts.hh

@@ -19,6 +19,7 @@ const static std::string userRegex = "(?:(?:" + unreservedRegex + "|" + pctEncod
 const static std::string authorityRegex = "(?:" + userRegex + "@)?" + hostRegex + "(?::[0-9]+)?";
 const static std::string pcharRegex = "(?:" + unreservedRegex + "|" + pctEncoded + "|" + subdelimsRegex + "|[:@])";
 const static std::string queryRegex = "(?:" + pcharRegex + "|[/? \"])*";
+const static std::string fragmentRegex = "(?:" + pcharRegex + "|[/? \"^])*";
 const static std::string segmentRegex = "(?:" + pcharRegex + "*)";
 const static std::string absPathRegex = "(?:(?:/" + segmentRegex + ")*/?)";
 const static std::string pathRegex = "(?:" + segmentRegex + "(?:/" + segmentRegex + ")*/?)";

src/libutil/url.cc

@@ -16,7 +16,7 @@ ParsedURL parseURL(const std::string & url)
         "((" + schemeRegex + "):"
         + "(?:(?://(" + authorityRegex + ")(" + absPathRegex + "))|(/?" + pathRegex + ")))"
         + "(?:\\?(" + queryRegex + "))?"
-        + "(?:#(" + queryRegex + "))?",
+        + "(?:#(" + fragmentRegex + "))?",
         std::regex::ECMAScript);
 
     std::smatch match;

tests/unit/libutil/url-name.cc

@@ -0,0 +1,67 @@
+#include "url-name.hh"
+#include <gtest/gtest.h>
+
+namespace nix {
+
+/* ----------- tests for url-name.hh --------------------------------------------------*/
+
+    TEST(getNameFromURL, getsNameFromURL) {
+        ASSERT_EQ(getNameFromURL(parseURL("path:/home/user/project")), "project");
+        ASSERT_EQ(getNameFromURL(parseURL("path:~/repos/nixpkgs#packages.x86_64-linux.hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("path:.#nonStandardAttr.mylaptop")), "nonStandardAttr.mylaptop");
+        ASSERT_EQ(getNameFromURL(parseURL("path:./repos/myflake#nonStandardAttr.mylaptop")), "nonStandardAttr.mylaptop");
+        ASSERT_EQ(getNameFromURL(parseURL("path:./nixpkgs#packages.x86_64-linux.complex^bin,man")), "complex");
+        ASSERT_EQ(getNameFromURL(parseURL("path:./myproj#packages.x86_64-linux.default^*")), "myproj");
+
+        ASSERT_EQ(getNameFromURL(parseURL("github:NixOS/nixpkgs#packages.x86_64-linux.hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("github:NixOS/nixpkgs#hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("github:NixOS/nix#packages.x86_64-linux.default")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("github:NixOS/nix#")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("github:NixOS/nix")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("github:cachix/devenv/main#packages.x86_64-linux.default")), "devenv");
+        ASSERT_EQ(getNameFromURL(parseURL("github:edolstra/nix-warez?rev=1234&dir=blender&ref=master")), "blender");
+
+        ASSERT_EQ(getNameFromURL(parseURL("gitlab:NixOS/nixpkgs#packages.x86_64-linux.hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("gitlab:NixOS/nixpkgs#hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("gitlab:NixOS/nix#packages.x86_64-linux.default")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("gitlab:NixOS/nix#")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("gitlab:NixOS/nix")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("gitlab:cachix/devenv/main#packages.x86_64-linux.default")), "devenv");
+
+        ASSERT_EQ(getNameFromURL(parseURL("sourcehut:NixOS/nixpkgs#packages.x86_64-linux.hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("sourcehut:NixOS/nixpkgs#hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("sourcehut:NixOS/nix#packages.x86_64-linux.default")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("sourcehut:NixOS/nix#")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("sourcehut:NixOS/nix")), "nix");
+        ASSERT_EQ(getNameFromURL(parseURL("sourcehut:cachix/devenv/main#packages.x86_64-linux.default")), "devenv");
+
+        ASSERT_EQ(getNameFromURL(parseURL("git://github.com/edolstra/dwarffs")), "dwarffs");
+        ASSERT_EQ(getNameFromURL(parseURL("git://github.com/edolstra/nix-warez?dir=blender")), "blender");
+        ASSERT_EQ(getNameFromURL(parseURL("git+file:///home/user/project")), "project");
+        ASSERT_EQ(getNameFromURL(parseURL("git+file:///home/user/project?ref=fa1e2d23a22")), "project");
+        ASSERT_EQ(getNameFromURL(parseURL("git+ssh://[email protected]/someuser/my-repo#")), "my-repo");
+        ASSERT_EQ(getNameFromURL(parseURL("git+git://github.com/someuser/my-repo?rev=v1.2.3")), "my-repo");
+        ASSERT_EQ(getNameFromURL(parseURL("git+ssh:///home/user/project?dir=subproject&rev=v2.4")), "subproject");
+        ASSERT_EQ(getNameFromURL(parseURL("git+http://not-even-real#packages.x86_64-linux.hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("git+https://not-even-real#packages.aarch64-darwin.hello")), "hello");
+
+        ASSERT_EQ(getNameFromURL(parseURL("tarball+http://github.com/NixOS/nix/archive/refs/tags/2.18.1#packages.x86_64-linux.jq")), "jq");
+        ASSERT_EQ(getNameFromURL(parseURL("tarball+https://github.com/NixOS/nix/archive/refs/tags/2.18.1#packages.x86_64-linux.hg")), "hg");
+        ASSERT_EQ(getNameFromURL(parseURL("tarball+file:///home/user/Downloads/nixpkgs-2.18.1#packages.aarch64-darwin.ripgrep")), "ripgrep");
+
+        ASSERT_EQ(getNameFromURL(parseURL("https://github.com/NixOS/nix/archive/refs/tags/2.18.1.tar.gz#packages.x86_64-linux.pv")), "pv");
+        ASSERT_EQ(getNameFromURL(parseURL("http://github.com/NixOS/nix/archive/refs/tags/2.18.1.tar.gz#packages.x86_64-linux.pv")), "pv");
+
+        ASSERT_EQ(getNameFromURL(parseURL("file:///home/user/project?ref=fa1e2d23a22")), "project");
+        ASSERT_EQ(getNameFromURL(parseURL("file+file:///home/user/project?ref=fa1e2d23a22")), "project");
+        ASSERT_EQ(getNameFromURL(parseURL("file+http://not-even-real#packages.x86_64-linux.hello")), "hello");
+        ASSERT_EQ(getNameFromURL(parseURL("file+http://gitfantasy.com/org/user/notaflake")), "notaflake");
+        ASSERT_EQ(getNameFromURL(parseURL("file+https://not-even-real#packages.aarch64-darwin.hello")), "hello");
+
+        ASSERT_EQ(getNameFromURL(parseURL("https://www.github.com/")), std::nullopt);
+        ASSERT_EQ(getNameFromURL(parseURL("path:.")), std::nullopt);
+        ASSERT_EQ(getNameFromURL(parseURL("file:.#")), std::nullopt);
+        ASSERT_EQ(getNameFromURL(parseURL("path:.#packages.x86_64-linux.default")), std::nullopt);
+        ASSERT_EQ(getNameFromURL(parseURL("path:.#packages.x86_64-linux.default^*")), std::nullopt);
+    }
+}

tests/unit/meson.build

@@ -52,6 +52,7 @@ libutil_tests_sources = files(
   'libutil/suggestions.cc',
   'libutil/tests.cc',
   'libutil/url.cc',
+  'libutil/url-name.cc',
   'libutil/xml-writer.cc',
 )