diple/ai_diff.go

package main

import (
	"bufio"
	"context"
	"fmt"
	"io"
	"net/http"
	"net/url"
	"path"
	"path/filepath"
	"strconv"
	"strings"
)

func (c *GitHubClient) PullRequestDiff(ctx context.Context, owner, repo string, number int) (string, error) {
	base := strings.TrimSuffix(c.endpoint, "/")
	switch {
	case base == "https://api.github.com/graphql":
		base = "https://api.github.com"
	case strings.HasSuffix(base, "/api/graphql"):
		base = strings.TrimSuffix(base, "/api/graphql") + "/api/v3"
	default:
		base = strings.TrimSuffix(base, "/graphql")
	}
	requestURL := base + "/repos/" + url.PathEscape(owner) + "/" + url.PathEscape(repo) +
		"/pulls/" + strconv.Itoa(number)
	req, err := http.NewRequestWithContext(ctx, http.MethodGet, requestURL, nil)
	if err != nil {
		return "", err
	}
	req.Header.Set("Authorization", "Bearer "+c.token)
	req.Header.Set("Accept", "application/vnd.github.v3.diff")
	req.Header.Set("User-Agent", "diple")
	response, err := c.http.Do(req)
	if err != nil {
		return "", err
	}
	defer response.Body.Close()
	if response.StatusCode != http.StatusOK {
		body, _ := io.ReadAll(io.LimitReader(response.Body, 4096))
		return "", fmt.Errorf("GitHub returned %s: %s", response.Status, strings.TrimSpace(string(body)))
	}
	body, err := io.ReadAll(io.LimitReader(response.Body, (32<<20)+1))
	if err != nil {
		return "", err
	}
	if len(body) > 32<<20 {
		return "", fmt.Errorf("PR diff exceeds the 32 MiB safety limit")
	}
	return string(body), nil
}

func (c *CachedGitHubService) PullRequestDiff(ctx context.Context, owner, repo string, number int) (string, error) {
	service, ok := c.remote.(AIDiffService)
	if !ok {
		return "", fmt.Errorf("configured GitHub service cannot load pull request diffs")
	}
	return service.PullRequestDiff(ctx, owner, repo, number)
}

func prepareAIDiff(raw string, config AIConfig) ([]aiDiffFile, []string, int) {
	sections := strings.Split(raw, "\ndiff --git ")
	var files []aiDiffFile
	var excluded []string
	redactions := 0
	for sectionIndex, section := range sections {
		if sectionIndex > 0 {
			section = "diff --git " + section
		}
		file, ok := parseAIDiffFile(section)
		if !ok {
			continue
		}
		reason := aiExcludedReason(file, config)
		if reason != "" {
			excluded = append(excluded, file.Path+" ("+reason+")")
			continue
		}
		redacted, count := redactAISecrets(file.Text)
		file.Text = sanitizeAIControls(redacted)
		redactions += count
		files = append(files, file)
	}
	return files, excluded, redactions
}

func parseAIDiffFile(section string) (aiDiffFile, bool) {
	scanner := bufio.NewScanner(strings.NewReader(section))
	scanner.Buffer(make([]byte, 4096), 2<<20)
	file := aiDiffFile{
		ChangedLines: make(map[int]bool), DeletedLines: make(map[int]bool), Text: section,
	}
	oldPath := ""
	oldLine := 0
	newLine := 0
	inHunk := false
	for scanner.Scan() {
		line := scanner.Text()
		if strings.HasPrefix(line, "--- ") && !inHunk {
			candidate := strings.TrimPrefix(line, "--- ")
			if strings.HasPrefix(candidate, `"`) {
				if unquoted, err := strconv.Unquote(candidate); err == nil {
					candidate = unquoted
				}
			}
			if strings.HasPrefix(candidate, "a/") {
				oldPath = strings.TrimPrefix(candidate, "a/")
			}
		}
		if strings.HasPrefix(line, "+++ ") && !inHunk {
			candidate := strings.TrimPrefix(line, "+++ ")
			if strings.HasPrefix(candidate, `"`) {
				if unquoted, err := strconv.Unquote(candidate); err == nil {
					candidate = unquoted
				}
			}
			if strings.HasPrefix(candidate, "b/") {
				file.Path = strings.TrimPrefix(candidate, "b/")
			}
		}
		if strings.HasPrefix(line, "@@ ") {
			parts := strings.Fields(line)
			if len(parts) >= 3 {
				oldRange := strings.TrimPrefix(parts[1], "-")
				oldStart := strings.SplitN(oldRange, ",", 2)[0]
				oldLine, _ = strconv.Atoi(oldStart)
				rangePart := strings.TrimPrefix(parts[2], "+")
				start := strings.SplitN(rangePart, ",", 2)[0]
				newLine, _ = strconv.Atoi(start)
				inHunk = true
			}
			continue
		}
		if !inHunk || line == "" {
			continue
		}
		switch line[0] {
		case '+':
			file.ChangedLines[newLine] = true
			newLine++
		case '-':
			file.DeletedLines[oldLine] = true
			oldLine++
		default:
			oldLine++
			newLine++
		}
	}
	if file.Path == "" {
		file.Path = oldPath
	}
	if file.Path == "" || file.Path == "/dev/null" {
		return aiDiffFile{}, false
	}
	file.Path = filepath.ToSlash(filepath.Clean(file.Path))
	if strings.HasPrefix(file.Path, "../") || filepath.IsAbs(file.Path) {
		return aiDiffFile{}, false
	}
	return file, len(file.ChangedLines) > 0 || len(file.DeletedLines) > 0
}

func aiExcludedReason(file aiDiffFile, config AIConfig) string {
	if len(file.Text) > config.MaxFileBytes {
		return "oversized"
	}
	if strings.Contains(file.Text, "GIT binary patch") ||
		strings.Contains(file.Text, "Binary files ") || strings.IndexByte(file.Text, 0) >= 0 {
		return "binary"
	}
	lower := strings.ToLower(file.Path)
	for _, pattern := range config.Exclude {
		pattern = filepath.ToSlash(pattern)
		if strings.HasSuffix(pattern, "/") {
			directory := strings.ToLower(pattern)
			if strings.HasPrefix(lower, directory) || strings.Contains(lower, "/"+directory) {
				return "excluded"
			}
		}
		if matched, _ := path.Match(strings.ToLower(pattern), path.Base(lower)); matched {
			return "excluded"
		}
		if matched, _ := path.Match(strings.ToLower(pattern), lower); matched {
			return "excluded"
		}
	}
	return ""
}