fix(pdf-compress): remove /screen preset that destroys font encoding
The -dPDFSETTINGS=/screen GS preset overwrites font encoding tables, producing garbled text in output PDFs. Replace with individual params that ONLY compress images while preserving fonts intact. Three quality levels via GS (no Stirling dependency): - extreme: 100 DPI, QFactor 1.2 (~quality 35) - high: 150 DPI, QFactor 0.76 (~quality 50) - balanced: 200 DPI, QFactor 0.4 (~quality 70) Route all UI modes through the GS endpoint with level parameter. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
AI Assistant
@@ -8,58 +8,75 @@ import { tmpdir } from "os";
|
||||
|
||||
const execFileAsync = promisify(execFile);
|
||||
|
||||
// Ghostscript args for extreme compression
|
||||
// Key: -dPassThroughJPEGImages=false forces recompression of existing JPEGs
|
||||
// QFactor 1.5 ≈ JPEG quality 25-30, matching iLovePDF extreme
|
||||
function gsArgs(input: string, output: string): string[] {
|
||||
// Ghostscript args for PDF compression.
|
||||
//
|
||||
// CRITICAL: Do NOT use -dPDFSETTINGS=/screen — it overrides font encoding
|
||||
// and produces garbled text. Instead, set each parameter individually so we
|
||||
// only compress IMAGES while keeping fonts and text intact.
|
||||
//
|
||||
// Strategy: recompress all raster images to JPEG at quality ~40-50,
|
||||
// downsample to 150 DPI, deduplicate, compress streams. Fonts untouched.
|
||||
function gsArgs(
|
||||
input: string,
|
||||
output: string,
|
||||
level: "extreme" | "high" | "balanced",
|
||||
): string[] {
|
||||
// Quality presets — only affect images, never fonts
|
||||
const presets = {
|
||||
extreme: { dpi: 100, qfactor: 1.2 }, // ~quality 35, aggressive
|
||||
high: { dpi: 150, qfactor: 0.76 }, // ~quality 50, good balance
|
||||
balanced: { dpi: 200, qfactor: 0.4 }, // ~quality 70, minimal loss
|
||||
};
|
||||
const { dpi, qfactor } = presets[level];
|
||||
|
||||
return [
|
||||
"-sDEVICE=pdfwrite",
|
||||
"-dCompatibilityLevel=1.5",
|
||||
"-dNOPAUSE",
|
||||
"-dBATCH",
|
||||
`-sOutputFile=${output}`,
|
||||
"-dPDFSETTINGS=/screen",
|
||||
// Force recompression of ALL images (the #1 key to matching iLovePDF)
|
||||
|
||||
// ── Image recompression (the main size reducer) ──
|
||||
// Force re-encode of existing JPEGs — without this, GS passes them through
|
||||
"-dPassThroughJPEGImages=false",
|
||||
"-dPassThroughJPXImages=false",
|
||||
// Use DCT (JPEG) for all color/gray images
|
||||
"-dAutoFilterColorImages=false",
|
||||
"-dAutoFilterGrayImages=false",
|
||||
"-dColorImageFilter=/DCTEncode",
|
||||
"-dGrayImageFilter=/DCTEncode",
|
||||
// Aggressive downsampling
|
||||
"-dEncodeColorImages=true",
|
||||
"-dEncodeGrayImages=true",
|
||||
|
||||
// ── Downsampling ──
|
||||
"-dDownsampleColorImages=true",
|
||||
"-dDownsampleGrayImages=true",
|
||||
"-dDownsampleMonoImages=true",
|
||||
"-dColorImageResolution=72",
|
||||
"-dGrayImageResolution=72",
|
||||
"-dMonoImageResolution=150",
|
||||
`-dColorImageResolution=${dpi}`,
|
||||
`-dGrayImageResolution=${dpi}`,
|
||||
`-dMonoImageResolution=${Math.max(dpi, 200)}`, // mono needs higher DPI
|
||||
"-dColorImageDownsampleType=/Bicubic",
|
||||
"-dGrayImageDownsampleType=/Bicubic",
|
||||
"-dColorImageDownsampleThreshold=1.0",
|
||||
"-dGrayImageDownsampleThreshold=1.0",
|
||||
"-dMonoImageDownsampleThreshold=1.0",
|
||||
// Encoding
|
||||
"-dEncodeColorImages=true",
|
||||
"-dEncodeGrayImages=true",
|
||||
// Font & structure
|
||||
"-dSubsetFonts=true",
|
||||
"-dEmbedAllFonts=true",
|
||||
|
||||
// ── Font handling — PRESERVE everything ──
|
||||
"-dSubsetFonts=true", // subset is safe — keeps encoding, reduces size
|
||||
"-dEmbedAllFonts=true", // ensure all fonts stay embedded
|
||||
"-dCompressFonts=true",
|
||||
|
||||
// ── Structure / stream optimization ──
|
||||
"-dCompressStreams=true",
|
||||
// CMYK→RGB (saves ~25% on CMYK images)
|
||||
"-sColorConversionStrategy=RGB",
|
||||
// Structure optimization
|
||||
"-dDetectDuplicateImages=true",
|
||||
"-dWriteXRefStm=true",
|
||||
"-dWriteObjStms=true",
|
||||
"-dPreserveMarkedContent=false",
|
||||
"-dOmitXMP=true",
|
||||
// JPEG quality dictionaries (QFactor 1.5 ≈ quality 25-30)
|
||||
"-sColorConversionStrategy=RGB", // CMYK→RGB saves ~25% on CMYK images
|
||||
|
||||
// ── JPEG quality dictionaries ──
|
||||
"-c",
|
||||
"<< /ColorACSImageDict << /QFactor 1.5 /Blend 1 /ColorTransform 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams",
|
||||
"<< /GrayACSImageDict << /QFactor 1.5 /Blend 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams",
|
||||
"<< /ColorImageDict << /QFactor 1.5 /Blend 1 /ColorTransform 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams",
|
||||
"<< /GrayImageDict << /QFactor 1.5 /Blend 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams",
|
||||
`<< /ColorACSImageDict << /QFactor ${qfactor} /Blend 1 /ColorTransform 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams`,
|
||||
`<< /GrayACSImageDict << /QFactor ${qfactor} /Blend 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams`,
|
||||
`<< /ColorImageDict << /QFactor ${qfactor} /Blend 1 /ColorTransform 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams`,
|
||||
`<< /GrayImageDict << /QFactor ${qfactor} /Blend 1 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> >> setdistillerparams`,
|
||||
"-f",
|
||||
input,
|
||||
];
|
||||
@@ -129,6 +146,42 @@ function extractFileFromMultipart(
|
||||
return null;
|
||||
}
|
||||
|
||||
/**
|
||||
* Extract a simple text field value from a multipart body.
|
||||
* Returns null if the field is not found.
|
||||
*/
|
||||
function extractFieldFromMultipart(
|
||||
raw: Buffer,
|
||||
boundary: string,
|
||||
fieldName: string,
|
||||
): string | null {
|
||||
const boundaryBuf = Buffer.from(`--${boundary}`);
|
||||
const headerSep = Buffer.from("\r\n\r\n");
|
||||
const crlf = Buffer.from("\r\n");
|
||||
const namePattern = `name="${fieldName}"`;
|
||||
|
||||
let searchFrom = 0;
|
||||
while (searchFrom < raw.length) {
|
||||
const partStart = raw.indexOf(boundaryBuf, searchFrom);
|
||||
if (partStart === -1) break;
|
||||
const lineEnd = raw.indexOf(crlf, partStart);
|
||||
if (lineEnd === -1) break;
|
||||
const headerEnd = raw.indexOf(headerSep, lineEnd);
|
||||
if (headerEnd === -1) break;
|
||||
|
||||
const headers = raw.subarray(lineEnd + 2, headerEnd).toString("utf8");
|
||||
if (headers.includes(namePattern) && !headers.includes("filename=")) {
|
||||
const valueStart = headerEnd + 4;
|
||||
const nextBoundary = raw.indexOf(Buffer.from(`\r\n--${boundary}`), valueStart);
|
||||
if (nextBoundary > valueStart) {
|
||||
return raw.subarray(valueStart, nextBoundary).toString("utf8").trim();
|
||||
}
|
||||
}
|
||||
searchFrom = headerEnd + 4;
|
||||
}
|
||||
return null;
|
||||
}
|
||||
|
||||
async function cleanup(dir: string) {
|
||||
try {
|
||||
const { readdir } = await import("fs/promises");
|
||||
@@ -186,14 +239,20 @@ export async function POST(req: NextRequest) {
|
||||
);
|
||||
}
|
||||
|
||||
// Extract compression level from multipart (optional "level" field)
|
||||
const levelParam = extractFieldFromMultipart(rawBuf, boundary, "level");
|
||||
const level: "extreme" | "high" | "balanced" =
|
||||
levelParam === "high" ? "high" :
|
||||
levelParam === "balanced" ? "balanced" : "extreme";
|
||||
|
||||
await writeFile(inputPath, pdfBuffer);
|
||||
const originalSize = pdfBuffer.length;
|
||||
|
||||
// Step 1: Ghostscript — aggressive image recompression + downsampling
|
||||
// Step 1: Ghostscript — image recompression + downsampling (fonts untouched)
|
||||
try {
|
||||
const { stderr } = await execFileAsync(
|
||||
"gs",
|
||||
gsArgs(inputPath, gsOutputPath),
|
||||
gsArgs(inputPath, gsOutputPath, level),
|
||||
{
|
||||
timeout: 300_000, // 5 min for very large files
|
||||
maxBuffer: 10 * 1024 * 1024, // 10MB stderr buffer
|
||||
|
||||
@@ -1662,15 +1662,10 @@ function PdfReducer() {
|
||||
try {
|
||||
const formData = new FormData();
|
||||
formData.append("fileInput", file);
|
||||
// All modes use the GS endpoint with a level parameter
|
||||
formData.append("level", mode === "extreme" ? "extreme" : mode === "max" ? "high" : "balanced");
|
||||
|
||||
let endpoint = "/api/compress-pdf";
|
||||
if (mode === "extreme") {
|
||||
endpoint = "/api/compress-pdf/extreme";
|
||||
} else {
|
||||
formData.append("optimizeLevel", mode === "max" ? "4" : "2");
|
||||
}
|
||||
|
||||
const res = await fetch(endpoint, {
|
||||
const res = await fetch("/api/compress-pdf/extreme", {
|
||||
method: "POST",
|
||||
body: formData,
|
||||
});
|
||||
@@ -1765,17 +1760,18 @@ function PdfReducer() {
|
||||
className="mt-1 w-full rounded-md border bg-background px-3 py-2 text-sm"
|
||||
>
|
||||
<option value="extreme">
|
||||
Compresie extremă — multi-pass (recomandat)
|
||||
Extremă — imagini 100 DPI, calitate scăzută
|
||||
</option>
|
||||
<option value="max">Compresie maximă — single-pass</option>
|
||||
<option value="balanced">Echilibrat</option>
|
||||
<option value="max">Puternică — imagini 150 DPI, calitate medie (recomandat)</option>
|
||||
<option value="balanced">Echilibrată — imagini 200 DPI, calitate bună</option>
|
||||
</select>
|
||||
{mode === "extreme" && (
|
||||
<p className="text-xs text-muted-foreground">
|
||||
Aplică compresie maximă în mai multe treceri succesive. Durează mai
|
||||
mult dar reduce semnificativ dimensiunea.
|
||||
</p>
|
||||
)}
|
||||
<p className="text-xs text-muted-foreground">
|
||||
{mode === "extreme"
|
||||
? "Reduce maxim dimensiunea. Imaginile pot pierde detalii fine."
|
||||
: mode === "max"
|
||||
? "Balanță bună între dimensiune și calitate. Recomandat pentru majoritatea fișierelor."
|
||||
: "Pierdere minimă de calitate. Ideal pentru documente cu grafice detaliate."}
|
||||
</p>
|
||||
</div>
|
||||
|
||||
<Button onClick={handleCompress} disabled={!file || loading}>
|
||||
|
||||
Reference in New Issue
Block a user