Skip to content
PDF Oxide

Getting Started with PDF Oxide (Objective-C)

PDF Oxide ships idiomatic Objective-C bindings over its Rust core — 0.8ms mean text extraction, 100% pass rate on 3,830 PDFs. NSObject wrappers (POXDocument, POXPdf) own the native handles and free them under ARC, returned strings come back as NSString, and any C-ABI error code surfaces as an NSError in the POXErrorDomain. New in v0.3.69.

Installation

The Objective-C binding links the default-feature cdylib and builds with clang under ARC. Build the native library, then make build against it:

# 1. build the native library (shipped binding feature set)
cargo build --release --lib --features ocr,rendering,signatures,barcodes,tsa-client,system-fonts

# 2. build the Objective-C binding (clang, ARC)
cd objc
make build PDF_OXIDE_LIB_DIR="$PWD/../target/release"
DYLD_LIBRARY_PATH="$PWD/../target/release" ./basic_extraction

Import the single public header in your sources:

#import "POXPdfOxide.h"

Quick Start

Open a PDF, inspect its metadata, and extract text from the first page. Every fallible call takes a trailing NSError**.

#import "POXPdfOxide.h"

NSError *err = nil;
POXDocument *doc = [POXDocument openPath:@"research-paper.pdf" error:&err];
if (!doc) {
    NSLog(@"open failed: %@", err.localizedDescription);
    return;
}

NSInteger pages = [doc pageCountError:&err];
POXVersion ver  = [doc version];
NSLog(@"pages: %ld  version: %d.%d", (long)pages, ver.major, ver.minor);

NSString *text = [doc extractText:0 error:&err];
NSLog(@"%@", text);

You can also open from in-memory bytes — useful when the PDF arrives over the network or from a database — and open password-protected files:

POXDocument *doc = [POXDocument openFromBytes:pdfData error:&err];

// Encrypted document, password supplied up front:
POXDocument *enc = [POXDocument openWithPassword:@"confidential.pdf"
                                       password:@"secret"
                                          error:&err];

// Or authenticate after opening:
BOOL ok = [doc authenticate:@"secret" error:&err];

Text Extraction

Plain text is the fast path. Extract a single page by its zero-based index, or pull the whole document at once.

// One page
NSString *text = [doc extractText:0 error:&err];

// Whole document, joined
NSString *all = [doc toPlainTextAllWithError:&err];

// Page by page
NSInteger count = [doc pageCountError:&err];
for (NSInteger i = 0; i < count; i++) {
    NSLog(@"--- page %ld ---\n%@", (long)i, [doc extractText:i error:&err]);
}

Words and Lines

extractWords: and extractTextLines: return arrays of element objects with bounding boxes and font metadata, all in PDF user-space points.

NSArray<POXWord *> *words = [doc extractWords:0 error:&err];
for (POXWord *w in words) {
    POXBbox box = w.bbox;
    NSLog(@"'%@' at (%.1f, %.1f) %.1fx%.1f  font=%@ size=%.1f  bold=%d",
          w.text, box.x, box.y, box.width, box.height,
          w.fontName, w.fontSize, w.bold);
}

NSArray<POXTextLine *> *lines = [doc extractTextLines:0 error:&err];
for (POXTextLine *line in lines) {
    NSLog(@"%@  (%ld words)", line.text, (long)line.wordCount);
}

POXChar (from extractChars:) exposes the same shape at character granularity — character, bbox, fontName, and fontSize.

Markdown and HTML

Convert a page — or the entire document — to Markdown or HTML.

// Single page
NSString *md   = [doc toMarkdown:0 error:&err];
NSString *html = [doc toHtml:0 error:&err];

// Whole document
NSString *mdAll   = [doc toMarkdownAllWithError:&err];
NSString *htmlAll = [doc toHtmlAllWithError:&err];

Search a single page with search:term:caseSensitive:error:, or the whole document with searchAll:caseSensitive:error:. Both return arrays of POXSearchResult carrying the matched text, page index, and bounding box.

NSArray<POXSearchResult *> *hits =
    [doc searchAll:@"configuration" caseSensitive:NO error:&err];

for (POXSearchResult *r in hits) {
    POXBbox b = r.bbox;
    NSLog(@"page %ld: '%@' at (%.0f, %.0f)", (long)r.page, r.text, b.x, b.y);
}

// Single-page variant:
NSArray<POXSearchResult *> *pageHits =
    [doc search:0 term:@"configuration" caseSensitive:NO error:&err];

Creating PDFs

The POXPdf builder produces PDFs from Markdown, HTML, or plain text. Save to a path or get the bytes back as NSData.

POXPdf *pdf = [POXPdf fromMarkdown:@"# Hello World\n\nThis is a PDF.\n"
                            error:&err];
[pdf saveToPath:@"output.pdf" error:&err];

// Or keep the bytes in memory
NSData *bytes = [pdf toBytesWithError:&err];

// HTML and plain text constructors exist too
POXPdf *invoice = [POXPdf fromHtml:@"<h1>Invoice</h1><p>Amount: $42</p>"
                            error:&err];
POXPdf *notes   = [POXPdf fromText:@"Plain text content." error:&err];

Round-trip from a builder straight into a document for extraction:

POXPdf *pdf      = [POXPdf fromMarkdown:@"# Report\n\nBody text.\n" error:&err];
POXDocument *doc = [POXDocument openFromBytes:[pdf toBytesWithError:&err]
                                        error:&err];
NSLog(@"%@", [doc extractText:0 error:&err]);

Error Handling

Every fallible method writes into the trailing NSError** and returns nil / a sentinel on failure. Errors land in the POXErrorDomain.

NSError *err = nil;
POXDocument *doc = [POXDocument openPath:@"document.pdf" error:&err];
if (!doc) {
    if ([err.domain isEqualToString:POXErrorDomain]) {
        NSLog(@"PDF error: %@", err.localizedDescription);
    }
    return;
}

NSString *text = [doc extractText:0 error:&err];
if (!text) {
    NSLog(@"extract failed: %@", err.localizedDescription);
}

Handles free themselves under ARC, but you can release the native handle eagerly with -close (idempotent):

[doc close];

Next Steps