runeandClaude Sonnet 5
69dbf17f5e
MCP's read_file previously hard-required UTF-8 text decoding, so any PDF in an allowed folder failed with "Cannot read file as UTF-8 text" — chat attachments already supported PDFs (raw bytes to vision-capable models), but the AI couldn't read one on its own during agentic file-tool use. search_files' content_search had the same gap, silently skipping PDFs. Adds MCPService.extractPDFText(atPath:) using PDFKit (built into macOS, no new dependency) to pull text from a PDF's text layer. Wired into both read_file and search_files' content search. Returns a clear error for scanned/image-only PDFs with no text layer. Automatically covers the Research Agents sub-agent tool loop too, since it shares the same executeTool dispatcher. Verified live: asked the AI to read a real PDF containing "The secret code is PINEAPPLE-42." via read_file — it extracted the text correctly through the MCP tool-call path. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
69 lines
2.5 KiB
Swift
69 lines
2.5 KiB
Swift
//
|
|
// MCPServicePDFTests.swift
|
|
// oAITests
|
|
//
|
|
// SPDX-License-Identifier: PolyForm-Noncommercial-1.0.0
|
|
// Copyright (C) 2026 Rune Olsen
|
|
|
|
import Testing
|
|
import Foundation
|
|
import CoreGraphics
|
|
import AppKit
|
|
@testable import oAI
|
|
|
|
@Suite("MCPService PDF text extraction")
|
|
struct MCPServicePDFTests {
|
|
|
|
/// Builds a real, well-formed single-page PDF (via CoreGraphics) containing the given text,
|
|
/// writes it to a temp file, and returns the path. Using CoreGraphics rather than hand-rolled
|
|
/// PDF bytes avoids flakiness from malformed xref tables etc.
|
|
private func makeTestPDF(text: String) throws -> String {
|
|
let path = FileManager.default.temporaryDirectory
|
|
.appendingPathComponent(UUID().uuidString)
|
|
.appendingPathExtension("pdf")
|
|
|
|
let pdfData = NSMutableData()
|
|
let consumer = CGDataConsumer(data: pdfData as CFMutableData)!
|
|
var mediaBox = CGRect(x: 0, y: 0, width: 200, height: 200)
|
|
let context = CGContext(consumer: consumer, mediaBox: &mediaBox, nil)!
|
|
|
|
context.beginPDFPage(nil)
|
|
let attributedString = NSAttributedString(string: text, attributes: [.font: NSFont.systemFont(ofSize: 24)])
|
|
let line = CTLineCreateWithAttributedString(attributedString)
|
|
context.textPosition = CGPoint(x: 10, y: 100)
|
|
CTLineDraw(line, context)
|
|
context.endPDFPage()
|
|
context.closePDF()
|
|
|
|
try (pdfData as Data).write(to: path)
|
|
return path.path
|
|
}
|
|
|
|
@Test("Extracts text from a real PDF's text layer")
|
|
func extractsTextFromRealPDF() throws {
|
|
let path = try makeTestPDF(text: "Hello World")
|
|
defer { try? FileManager.default.removeItem(atPath: path) }
|
|
|
|
let extracted = MCPService.shared.extractPDFText(atPath: path)
|
|
#expect(extracted?.contains("Hello World") == true)
|
|
}
|
|
|
|
@Test("Returns nil for a nonexistent path")
|
|
func returnsNilForMissingFile() {
|
|
let extracted = MCPService.shared.extractPDFText(atPath: "/tmp/definitely-does-not-exist-\(UUID().uuidString).pdf")
|
|
#expect(extracted == nil)
|
|
}
|
|
|
|
@Test("Returns nil for a file that isn't a valid PDF")
|
|
func returnsNilForGarbageBytes() throws {
|
|
let path = FileManager.default.temporaryDirectory
|
|
.appendingPathComponent(UUID().uuidString)
|
|
.appendingPathExtension("pdf")
|
|
try Data("not a real pdf".utf8).write(to: path)
|
|
defer { try? FileManager.default.removeItem(at: path) }
|
|
|
|
let extracted = MCPService.shared.extractPDFText(atPath: path.path)
|
|
#expect(extracted == nil)
|
|
}
|
|
}
|