sift/internal/pdftest/pdftest.go

// Package pdftest builds small, deterministic PDFs for tests: text pages in Helvetica, an optional
// outline (bookmarks), and image-only ("scanned") pages. It exists so the P5a tests need no binary
// fixtures from elsewhere. It is test support: nothing outside _test.go files imports it.
package pdftest

import (
	"bytes"
	"fmt"
	"strings"
)

// Outline is one bookmark: a title that points at a 1-based page, nested at Level (1 = top).
// Build supports one level; Level is kept so tests can state what they expect back.
type Outline struct {
	Title string
	Page  int
	Level int
}

// Doc describes a PDF to build. Pages holds each page's text; a page's lines are split on "\n"
// and set one per line (long lines are not wrapped: callers keep them under ~90 characters).
// Scanned pages (1-based) carry a small image and no text at all.
type Doc struct {
	Title   string
	Pages   []string
	Outline []Outline
	Scanned []int
	// NamedDests writes the outline the way LaTeX's hyperref and Acrobat do: each item is a
	// /GoTo action to a named destination (resolved through the catalog's /Names /Dests tree)
	// and its /Title is an indirect object. Without it items carry a direct /Dest array.
	NamedDests bool
}

// Build returns the PDF bytes for d. It panics on a Doc it cannot express (an outline entry
// pointing past the last page), since that is a bug in the test that built it.
func Build(d Doc) []byte {
	var b bytes.Buffer
	var offsets []int
	obj := func(body string) int {
		offsets = append(offsets, b.Len())
		n := len(offsets)
		fmt.Fprintf(&b, "%d 0 obj\n%s\nendobj\n", n, body)
		return n
	}
	b.WriteString("%PDF-1.4\n%\xe2\xe3\xcf\xd3\n")

	// Object numbers are fixed by position: 1 catalog, 2 pages, 3 font, 4 image, 5 info, then
	// two per page (page, content), then the outline.
	nPages := len(d.Pages)
	firstPage := 6
	pageObj := func(i int) int { return firstPage + 2*i }
	outlineRoot := firstPage + 2*nPages
	scanned := map[int]bool{}
	for _, p := range d.Scanned {
		scanned[p] = true
	}

	// With NamedDests the outline items are followed by one title object per item and then
	// the name tree, so the numbers below stay fixed by position.
	nameTree := outlineRoot + 1 + 2*len(d.Outline)
	catalog := "<< /Type /Catalog /Pages 2 0 R"
	if len(d.Outline) > 0 {
		catalog += fmt.Sprintf(" /Outlines %d 0 R /PageMode /UseOutlines", outlineRoot)
		if d.NamedDests {
			catalog += fmt.Sprintf(" /Names << /Dests %d 0 R >>", nameTree)
		}
	}
	obj(catalog + " >>")
	var kids []string
	for i := range d.Pages {
		kids = append(kids, fmt.Sprintf("%d 0 R", pageObj(i)))
	}
	obj(fmt.Sprintf("<< /Type /Pages /Kids [%s] /Count %d >>", strings.Join(kids, " "), nPages))
	obj("<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica /Encoding /WinAnsiEncoding >>")
	img := []byte{0x00, 0x80, 0x80, 0x00} // 2x2 grey pixels
	obj(fmt.Sprintf("<< /Type /XObject /Subtype /Image /Width 2 /Height 2 /ColorSpace /DeviceGray /BitsPerComponent 8 /Length %d >>\nstream\n%s\nendstream", len(img), img))
	obj(fmt.Sprintf("<< /Title (%s) /Producer (sift pdftest) >>", esc(d.Title)))

	for i, text := range d.Pages {
		var content string
		if scanned[i+1] {
			content = "q 400 0 0 500 100 150 cm /Im1 Do Q"
		} else {
			var c strings.Builder
			c.WriteString("BT /F1 11 Tf 13 TL 60 760 Td\n")
			for _, line := range strings.Split(text, "\n") {
				fmt.Fprintf(&c, "(%s) Tj T*\n", esc(line))
			}
			c.WriteString("ET")
			content = c.String()
		}
		obj(fmt.Sprintf("<< /Type /Page /Parent 2 0 R /MediaBox [0 0 612 792] /Resources << /Font << /F1 3 0 R >> /XObject << /Im1 4 0 R >> >> /Contents %d 0 R >>", pageObj(i)+1))
		obj(fmt.Sprintf("<< /Length %d >>\nstream\n%s\nendstream", len(content), content))
	}

	if len(d.Outline) > 0 {
		first := outlineRoot + 1
		last := outlineRoot + len(d.Outline)
		obj(fmt.Sprintf("<< /Type /Outlines /First %d 0 R /Last %d 0 R /Count %d >>", first, last, len(d.Outline)))
		for i, o := range d.Outline {
			if o.Page < 1 || o.Page > nPages {
				panic(fmt.Sprintf("pdftest: outline %q points at page %d of %d", o.Title, o.Page, nPages))
			}
			n := first + i
			item := fmt.Sprintf("<< /Title (%s) /Parent %d 0 R /Dest [%d 0 R /XYZ null null null]", esc(o.Title), outlineRoot, pageObj(o.Page-1))
			if d.NamedDests {
				item = fmt.Sprintf("<< /Title %d 0 R /Parent %d 0 R /A << /S /GoTo /D (dest.%d) >>", last+1+i, outlineRoot, i)
			}
			if i > 0 {
				item += fmt.Sprintf(" /Prev %d 0 R", n-1)
			}
			if n < last {
				item += fmt.Sprintf(" /Next %d 0 R", n+1)
			}
			obj(item + " >>")
		}
		if d.NamedDests {
			var names []string
			for i, o := range d.Outline {
				obj(fmt.Sprintf("(%s)", esc(o.Title)))
				names = append(names, fmt.Sprintf("(dest.%d) [%d 0 R /XYZ null null null]", i, pageObj(o.Page-1)))
			}
			obj(fmt.Sprintf("<< /Names [%s] >>", strings.Join(names, " ")))
		}
	}

	xref := b.Len()
	fmt.Fprintf(&b, "xref\n0 %d\n0000000000 65535 f \n", len(offsets)+1)
	for _, off := range offsets {
		fmt.Fprintf(&b, "%010d 00000 n \n", off)
	}
	fmt.Fprintf(&b, "trailer\n<< /Size %d /Root 1 0 R /Info 5 0 R >>\nstartxref\n%d\n%%%%EOF\n", len(offsets)+1, xref)
	return b.Bytes()
}

// Lorem returns n lines of distinct, searchable filler: "p<page> l<line> …", so a test can tell
// which page a line came from.
func Lorem(page, n int) string {
	words := []string{"alpha", "bravo", "charlie", "delta", "echo", "foxtrot", "golf", "hotel", "india", "juliet", "kilo", "lima"}
	var lines []string
	for i := 1; i <= n; i++ {
		var w []string
		for j := 0; j < 10; j++ {
			w = append(w, words[(page*7+i*3+j)%len(words)])
		}
		lines = append(lines, fmt.Sprintf("p%d l%d %s", page, i, strings.Join(w, " ")))
	}
	return strings.Join(lines, "\n")
}

func esc(s string) string {
	r := strings.NewReplacer(`\`, `\\`, `(`, `\(`, `)`, `\)`)
	return r.Replace(s)
}