Initial Commit

2025-02-12 19:33:11 -08:00
commit 76cbbad74a
43 changed files with 4970 additions and 0 deletions
--- a/internal/index/apt.go
+++ b/internal/index/apt.go
@@ -0,0 +1,114 @@
+/*
+ * distrohop - A utility for correlating and identifying equivalent software
+ * packages across different Linux distributions
+ *
+ * Copyright (C) 2025 Elara Ivy <elara@elara.ws>
+ *
+ * This file is part of distrohop.
+ *
+ * distrohop is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License, or (at your option) any later version.
+ *
+ * distrohop is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with distrohop.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+package index
+
+import (
+	"bufio"
+	"context"
+	"errors"
+	"io"
+	"net/url"
+	"strings"
+
+	"github.com/mholt/archives"
+	"go.elara.ws/distrohop/internal/tags"
+)
+
+type APT struct{}
+
+func (APT) Name() string {
+	return "apt"
+}
+
+func (APT) IndexURL(baseURL, version, repo, arch string) ([]string, error) {
+	indexURL, err := url.JoinPath(baseURL, "dists", version, repo, "Contents-"+arch+".gz")
+	if err != nil {
+		return nil, err
+	}
+	// Before Debian Wheezy, the path to Contents indices didn't include $COMP/repo, so we need to try
+	// both the new and old URL formats. Ubuntu also still uses the pre-Debian-Wheezy convention.
+	deprecatedURL, err := url.JoinPath(baseURL, "dists", version, "Contents-"+arch+".gz")
+	if err != nil {
+		return nil, err
+	}
+	return []string{indexURL, deprecatedURL}, nil
+}
+
+func (APT) ReadPkgData(r io.Reader, out chan Record) {
+	ctx := context.Background()
+	format, r, err := archives.Identify(ctx, "", r)
+	if err != nil {
+		out <- Record{Error: err}
+		return
+	}
+
+	decomp, ok := format.(archives.Decompressor)
+	if !ok {
+		out <- Record{Error: errors.New("downloaded index is not a valid compressed file")}
+		return
+	}
+
+	dr, err := decomp.OpenReader(r)
+	if err != nil {
+		out <- Record{Error: err}
+		return
+	}
+	defer dr.Close()
+
+	br := bufio.NewReader(dr)
+	for {
+		line, err := br.ReadString('\n')
+		if errors.Is(err, io.EOF) {
+			close(out)
+			break
+		} else if err != nil {
+			out <- Record{Error: err}
+			return
+		}
+
+		lastSpaceIdx := strings.LastIndexByte(line, ' ')
+		if lastSpaceIdx == -1 {
+			continue
+		}
+
+		fpath := "/" + strings.TrimSpace(line[:lastSpaceIdx])
+		names := strings.Split(strings.TrimSpace(line[lastSpaceIdx+1:]), ",")
+		for _, name := range names {
+			slashIdx := strings.LastIndexByte(name, '/')
+			if slashIdx != -1 {
+				name = name[slashIdx+1:]
+			}
+
+			if strings.Contains(fpath, "changelog.Debian") ||
+				strings.Contains(fpath, "README.Debian") ||
+				strings.Contains(fpath, "NEWS.Debian.gz") {
+				continue
+			}
+
+			out <- Record{
+				Name: name,
+				Tags: tags.Generate(fpath),
+			}
+		}
+	}
+}
--- a/internal/index/dnf.go
+++ b/internal/index/dnf.go
@@ -0,0 +1,154 @@
+/*
+ * distrohop - A utility for correlating and identifying equivalent software
+ * packages across different Linux distributions
+ *
+ * Copyright (C) 2025 Elara Ivy <elara@elara.ws>
+ *
+ * This file is part of distrohop.
+ *
+ * distrohop is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License, or (at your option) any later version.
+ *
+ * distrohop is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with distrohop.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+package index
+
+import (
+	"bufio"
+	"context"
+	"encoding/xml"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"strings"
+
+	"github.com/mholt/archives"
+	"go.elara.ws/distrohop/internal/tags"
+)
+
+type repomd struct {
+	Locations []location `xml:"data>location"`
+}
+
+type location struct {
+	Href string `xml:"href,attr"`
+}
+
+func (r repomd) getGzipFile() string {
+	for _, loc := range r.Locations {
+		if strings.HasSuffix(loc.Href, "filelists.xml.gz") {
+			return loc.Href
+		}
+	}
+	return "<unknown>"
+}
+
+type DNF struct{}
+
+func (DNF) Name() string {
+	return "dnf"
+}
+
+func (DNF) IndexURL(baseURL, version, repo, arch string) ([]string, error) {
+	u, err := url.ParseRequestURI(baseURL)
+	if err != nil {
+		return nil, err
+	}
+	repomdPath := fmt.Sprintf("/pub/fedora/linux/releases/%s/%s/%s/os/repodata/repomd.xml", version, repo, arch)
+	u.Path = repomdPath
+
+	res, err := http.Get(u.String())
+	if err != nil {
+		return nil, err
+	}
+	defer res.Body.Close()
+
+	var data repomd
+	err = xml.NewDecoder(res.Body).Decode(&data)
+	if err != nil {
+		return nil, err
+	}
+
+	gzipFile := data.getGzipFile()
+	if gzipFile == "" {
+		return nil, errors.New("no gzip file found in repomd.xml")
+	}
+
+	u.Path = fmt.Sprintf("/pub/fedora/linux/releases/%s/%s/%s/os/%s", version, repo, arch, gzipFile)
+	return []string{u.String()}, nil
+}
+
+func (DNF) ReadPkgData(r io.Reader, out chan Record) {
+	ctx := context.Background()
+	format, r, err := archives.Identify(ctx, "", r)
+	if err != nil {
+		out <- Record{Error: err}
+		return
+	}
+
+	decomp, ok := format.(archives.Decompressor)
+	if !ok {
+		out <- Record{Error: errors.New("downloaded index is not a valid compressed file")}
+		return
+	}
+
+	dr, err := decomp.OpenReader(r)
+	if err != nil {
+		out <- Record{Error: err}
+		return
+	}
+	defer dr.Close()
+
+	br := bufio.NewReader(dr)
+	var currentPkg string
+
+	for {
+		line, err := br.ReadString('\n')
+		if errors.Is(err, io.EOF) {
+			close(out)
+			break
+		} else if err != nil {
+			out <- Record{Error: err}
+			return
+		}
+		line = strings.TrimSpace(line)
+
+		switch {
+		case strings.HasPrefix(line, "<file"):
+			// Skip directories and symlinks
+			if strings.HasPrefix(line[5:], `type="dir"`) || line[5] == 'l' {
+				continue
+			}
+
+			start := strings.IndexByte(line, '>') + 1
+			end := strings.LastIndexByte(line, '<')
+			fpath := line[start:end]
+
+			if strings.Contains(fpath, ".build-id") {
+				continue
+			}
+
+			out <- Record{
+				Name: currentPkg,
+				Tags: tags.Generate(fpath),
+			}
+		case strings.HasPrefix(line, "<package"):
+			start := strings.LastIndex(line, `name="`) + 6
+			end := start + strings.IndexByte(line[start:], '"')
+			currentPkg = line[start:end]
+		default:
+			continue
+		}
+	}
+}
--- a/internal/index/index.go
+++ b/internal/index/index.go
@@ -0,0 +1,60 @@
+/*
+ * distrohop - A utility for correlating and identifying equivalent software
+ * packages across different Linux distributions
+ *
+ * Copyright (C) 2025 Elara Ivy <elara@elara.ws>
+ *
+ * This file is part of distrohop.
+ *
+ * distrohop is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License, or (at your option) any later version.
+ *
+ * distrohop is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with distrohop.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+package index
+
+import (
+	"fmt"
+	"io"
+)
+
+// Record represents a data record for a single package
+type Record struct {
+	Name  string
+	Tags  []string
+	Error error
+}
+
+type Importer interface {
+	// Name returns the name of the importer
+	Name() string
+	// IndexURL generates a list of possible index URLs to try
+	IndexURL(baseURL, version, repo, arch string) ([]string, error)
+	// ReadPkgData reads data from an index file and sends it on out
+	ReadPkgData(r io.Reader, out chan Record)
+}
+
+var importers = []Importer{
+	APT{},
+	DNF{},
+	Pacman{},
+}
+
+// GetImporter gets an importer by its name
+func GetImporter(name string) (Importer, error) {
+	for _, importer := range importers {
+		if importer.Name() == name {
+			return importer, nil
+		}
+	}
+	return nil, fmt.Errorf("no such importer: %q", name)
+}
--- a/internal/index/pacman.go
+++ b/internal/index/pacman.go
@@ -0,0 +1,145 @@
+/*
+ * distrohop - A utility for correlating and identifying equivalent software
+ * packages across different Linux distributions
+ *
+ * Copyright (C) 2025 Elara Ivy <elara@elara.ws>
+ *
+ * This file is part of distrohop.
+ *
+ * distrohop is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License, or (at your option) any later version.
+ *
+ * distrohop is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with distrohop.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+package index
+
+import (
+	"archive/tar"
+	"bufio"
+	"bytes"
+	"context"
+	"errors"
+	"io"
+	"net/url"
+	"os"
+	"path"
+	"strings"
+
+	"github.com/mholt/archives"
+	"go.elara.ws/distrohop/internal/tags"
+)
+
+type Pacman struct{}
+
+func (Pacman) Name() string {
+	return "pacman"
+}
+
+func (Pacman) IndexURL(baseURL, version, repo, arch string) ([]string, error) {
+	baseURL = os.Expand(baseURL, func(s string) string {
+		switch s {
+		case "repo":
+			return repo
+		case "arch":
+			return arch
+		}
+		return "$" + s
+	})
+
+	u, err := url.ParseRequestURI(baseURL)
+	if err != nil {
+		return nil, err
+	}
+	filePath, err := url.JoinPath(u.Path, repo+".files")
+	if err != nil {
+		return nil, err
+	}
+	u.Path = filePath
+	return []string{u.String()}, nil
+}
+
+func (Pacman) ReadPkgData(r io.Reader, out chan Record) {
+	ctx := context.Background()
+	format, r, err := archives.Identify(ctx, "", r)
+	if err != nil {
+		out <- Record{Error: err}
+		return
+	}
+
+	decomp, ok := format.(archives.Decompressor)
+	if !ok {
+		out <- Record{Error: errors.New("downloaded index is not a valid compressed file")}
+		return
+	}
+
+	dr, err := decomp.OpenReader(r)
+	if err != nil {
+		out <- Record{Error: err}
+		return
+	}
+	defer dr.Close()
+
+	tr := tar.NewReader(dr)
+	var currentPkg string
+
+	for {
+		hdr, err := tr.Next()
+		if errors.Is(err, io.EOF) {
+			close(out)
+			break
+		} else if err != nil {
+			out <- Record{Error: err}
+			return
+		}
+
+		switch path.Base(hdr.Name) {
+		case "desc":
+			data, err := io.ReadAll(tr)
+			if err != nil {
+				out <- Record{Error: err}
+				return
+			}
+
+			labelIdx := bytes.Index(data, []byte("%NAME%\n"))
+			if labelIdx == -1 {
+				continue
+			}
+
+			start := labelIdx + 7
+			end := start + bytes.IndexByte(data[start:], '\n')
+			currentPkg = string(data[start:end])
+		case "files":
+			br := bufio.NewReader(tr)
+			for {
+				fpath, err := br.ReadString('\n')
+				if errors.Is(err, io.EOF) {
+					break
+				} else if err != nil {
+					out <- Record{Error: err}
+					return
+				}
+
+				fpath = strings.TrimSpace(fpath)
+				if fpath == "%FILES%" || strings.HasSuffix(fpath, "/") {
+					continue
+				}
+
+				fpath = "/" + fpath
+
+				out <- Record{
+					Name: currentPkg,
+					Tags: tags.Generate(fpath),
+				}
+			}
+		}
+	}
+}