pdfrb 0.7.43 → 0.7.45

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: b0b004db58a66ef4268db7138eb13c2813fcceb4e22a492674597f631d6f889e
-  data.tar.gz: b29ee27e11e4aea41a482e3eb217c7ddf70a5cdf4b21f70b2d16be01747822df
+  metadata.gz: 9283c2555bc64be8606a06fb6fc62c9498382499c30c7eea8afd36c4dcfc6a82
+  data.tar.gz: e93caaff66db1dcf3944873ad5b41fe9a8876623d33ae0c53918461a6f0a4c49
 SHA512:
-  metadata.gz: 25dc0c66f65bf454f2ac91c362031c57720d060edaeb54912aba896115216fbe43ce4fd947e21e94a445e521168e1eb629d51db42dc38c9ca5639ca41e8786e5
-  data.tar.gz: 3368f119c878bb7bcce11f21a4bd6edb267e4605ae079ca85dafbce6defa1397e9b5f0649bbdfc3a7084d161eb0f6c20da934b5b0f288344538f296703a031c7
+  metadata.gz: 16b7e2394f0275d631d4ba0946e3b0d5f2973240f9fe1d4f5eebfbb4998e666e6e0ff9e62dd50fd43eebb5ecb7b418846d6ded5be052b1c3462f0d3033cbd1a3
+  data.tar.gz: 4219346da2b075bb87024963f248c1a8e26852f9e928c24f0155291e242bd8d4ff28175f74ef1719c05812fc34aa0782d4f81d2a226c108e996d5940d48ab6fe

data/CHANGELOG.md

@@ -95,6 +95,74 @@ All notable changes to the pdfrb gem will be documented in this file.
 * Removed all `instance_variable_set/get`, `send` to private
   methods, `respond_to?` for type checks.
 
+## [0.7.44] — 2026-08-26
+
+### Added
+
+* Top-level `Pdfrb.open(path)` / `Pdfrb.parse(bytes_or_io)` module
+  functions.
+* `Document#encrypt!` / `#decrypt!` delegates;
+  `ExtractText.call_single_page(page)`.
+* `Canvas#rectangle(point:, width:, height:)` and
+  `Canvas#line(from:, to:)` keyword forms alongside positional.
+* Executable cookbook spec (`usage_doc_spec.rb`): every docs/USAGE.md
+  snippet runs against temp files.
+
+### Fixed
+
+* **Oid allocation on parsed documents** started at 1 regardless of
+  the xref, so mutating a reopened PDF overwrote existing objects —
+  `Task::Merge` dropped the target's original pages (1+2 -> 2).
+  Allocation now seeds past max(xref oid, trailer /Size).
+
+## [0.7.43] — 2026-08-26
+
+### Added
+
+* `Content::InlineImage` value object for BI/ID/EI sequences:
+  geometry accessors, s8.9.7 abbreviation expansion (color spaces
+  and filter names), components/expected_decoded_size, and
+  `decoded_data` applying the filter pipeline on demand.
+
+### Fixed
+
+* operators/inline_image.rb called `register` inside `class << self`
+  (receiver = the singleton class) — never legitimately executable;
+  moved to class-body level.
+* operators.rb autoloaded `:InlineImage` from a file defining
+  `Operator::BeginInlineImage` (namespace mismatch); correct
+  BeginInlineImage/EndInlineImage autoloads now live in operator.rb.
+
+## [0.7.42] — 2026-08-25
+
+### Fixed
+
+* Packing-size spec no longer skips: rebuilt with 60 small
+  annotations so object-stream packing wins decisively (~6.5KB ->
+  ~1.4KB). Suite is now 0 failures, 0 pending.
+
+## [0.7.41] — 2026-08-25
+
+### Added
+
+* Arlington conformance fixture walk is real: all 12 Annex H
+  fixtures validate through the typed field validator with zero
+  violations (previously a pending stub).
+
+### Fixed
+
+* `PageTreeNodeRoot` shadows the inherited required `/Parent` (the
+  root TSV omits the key; the root must not carry a parent).
+* `Dictionary.each_field` honors field shadowing: a subclass field
+  with the same name fully replaces the inherited one.
+
+## [0.7.40] — 2026-08-25
+
+### Added
+
+* CHANGELOG backfill for 0.7.1 through 0.7.39 (29 releases
+  reconstructed from tag history).
+
 ## [0.7.39] — 2026-08-24
 
 ### Added

data/CLAUDE.md

@@ -105,6 +105,22 @@ These repos live outside the project tree but are the authoritative references.
 | `~/src/pdfa/technote-pdfua1-001/` | PDF/UA-1 Technical Note 001 — `ActualText` on `Figure` structure elements. |
 | `~/src/pdfa/technote-pdfua1-002/` | PDF/UA-1 Technical Note 002 — `Reference` structure elements containing links. |
 
-## Work in progress
+## Status
 
-The gem is at 0.1.0 — foundation only (autoload tree, error hierarchy, constants, configuration). All remaining implementation work is tracked as individual TODO files in `TODO.general-rels/`. The master index is `TODO.general-rels/README.md`. Pick up work in dependency order; each TODO file's front-matter declares `depends_on`.
+Released on RubyGems (0.7.x, BSD-2-Clause). The full stack is
+implemented: reader/writer round-trip, content streams + Canvas,
+encryption (RC4/AES-256), signatures, layout/Composer, conformance
+validation (PDF/A, PDF/UA, PDF/X, PDF/VT), and the CLI.
+
+**Arlington TSV coverage is complete** — 612 of 614 non-alias TSVs
+map to `Model::Type::*` classes (only the two single-row OPI alias
+wrappers remain, deliberately). Every commit goes through
+PR + CI; the CHANGELOG records each release.
+
+Known depth gaps: CFF/OTF font subsetting (only TrueType subsets),
+veraPDF cross-check for PDF/A output. The original plan files live
+in `TODO.general-rels/` (~157 entries, mostly landed).
+
+`docs/USAGE.md` is the verified cookbook — `spec/pdfrb/usage_doc_spec.rb`
+executes every snippet, so keep both in sync when changing public
+APIs.

data/README.adoc

@@ -31,7 +31,7 @@ https://github.com/gettalong/hexapdf[HexaPDF].
 gem install pdfrb
 ----
 
-== Usage (sketch — APIs land across the TODO sequence)
+== Usage
 
 [source,ruby]
 ----
@@ -39,15 +39,29 @@ require "pdfrb"
 
 # Read
 doc = Pdfrb.open("input.pdf")
-puts doc.catalog.pages.count
+puts doc.pages.count
+puts Pdfrb::Task::ExtractText.call_single_page(doc.pages.first)
 
 # Create
 doc = Pdfrb::Document.new
+font = doc.fonts.add("Helvetica")
 page = doc.pages.add
-page.canvas.text("Hello, PDF", at: [72, 720])
+page.canvas.text("Hello, PDF", at: [72, 720], font: font, size: 24)
+page.canvas.rectangle(point: [72, 700], width: 200, height: 50)
+page.canvas.stroke
 doc.write("hello.pdf")
+
+# Merge
+target = Pdfrb.open("base.pdf")
+Pdfrb::Task::Merge.call(target, Pdfrb.open("appendix.pdf"))
+target.write("merged.pdf")
 ----
 
+A full cookbook (tagged PDF, layers, forms, signatures, diff,
+conformance validation, linearization, encryption, CLI) lives in
+link:docs/USAGE.md[docs/USAGE.md] — every snippet there is executed
+by the test suite.
+
 == Architecture
 
 [cols="1,4", options="header"]

data/docs/USAGE.md

@@ -137,9 +137,18 @@ File.binwrite("linearized.pdf", io.string)
 ## Encryption
 
 ```ruby
-doc = Pdfrb.open("input.pdf")
+doc = Pdfrb::Document.new
+doc.pages.add
+doc.encrypt!(user_password: "secret", owner_password: "owner", bits: 128)
 doc.write("encrypted.pdf")
-# Encryption configuration via document.config
+
+# Opening an encrypted document:
+doc = Pdfrb.open("encrypted.pdf",
+                 config: { "encryption.password" => "secret" })
+
+# Removing encryption from a parsed document:
+doc.decrypt!
+doc.write("decrypted.pdf")
 ```
 
 ## Merging PDFs

data/lib/pdfrb/content/canvas.rb

@@ -64,7 +64,11 @@ module Pdfrb
         self
       end
 
-      def line(x1, y1, x2, y2)
+      def line(x1 = nil, y1 = nil, x2 = nil, y2 = nil, from: nil, to: nil)
+        if from && to
+          x1, y1 = from
+          x2, y2 = to
+        end
         move_to(x1, y1).line_to(x2, y2)
       end
 
@@ -189,8 +193,14 @@ module Pdfrb
         self
       end
 
-      def rectangle(x, y, width, height)
-        emit_op(Pdfrb::Content::Operator::Rectangle, x, y, width, height)
+      def rectangle(x = nil, y = nil, w = nil, h = nil,
+                    point: nil, width: nil, height: nil)
+        x, y = point if point
+        w = width if w.nil?
+        h = height if h.nil?
+        raise ArgumentError, "rectangle needs x, y, width and height" if [x, y, w, h].any?(&:nil?)
+
+        emit_op(Pdfrb::Content::Operator::Rectangle, x, y, w, h)
         self
       end
 

data/lib/pdfrb/document.rb

@@ -166,6 +166,14 @@ module Pdfrb
 
     def encryption; @encryption ||= Document::Encryption.new(self); end
 
+    def encrypt!(**)
+      encryption.encrypt!(**)
+    end
+
+    def decrypt!
+      encryption.decrypt!
+    end
+
     def info; @info ||= Document::Info.new(self); end
 
     def display; @display ||= Document::Display.new(self); end
@@ -306,6 +314,19 @@ module Pdfrb
 
       @xref, @trailer_dict = load_xref_and_trailer(io, sxref)
       @object_reader = Pdfrb::Source::ObjectReader.new(self, @xref) if @xref
+      seed_next_oid
+    end
+
+    # New/modified objects must not collide with objects loaded from
+    # the xref: continue allocating past the highest known oid (or
+    # the trailer /Size, whichever is larger).
+    def seed_next_oid
+      return if @xref.nil?
+
+      max_oid = @xref.entries.keys.max || 0
+      size = @trailer_dict && @trailer_dict[:Size]
+      size = size.to_i if size.respond_to?(:to_i)
+      @next_oid = [max_oid + 1, size.to_i, @next_oid].max
     end
 
     def load_xref_and_trailer(io, sxref)

data/lib/pdfrb/task/extract_text.rb

@@ -31,6 +31,18 @@ module Pdfrb
         block_given? ? document : results
       end
 
+      # Extract the text of a single page (see #call for the
+      # resolution strategy).
+      def call_single_page(page)
+        extractor = TextCollector.new(page.document, page)
+        begin
+          extractor.process(page.decoded_content)
+        rescue StandardError
+          # Malformed content stream — best effort, skip.
+        end
+        extractor.text
+      end
+
       # Internal: collects bytes from +Tj+ / +TJ+ / +'+ / +"'.
       # Tracks text-matrix y to insert line breaks at vertical moves.
       # Resolves glyph codes to Unicode via the active font's

data/lib/pdfrb/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Pdfrb
-  VERSION = "0.7.43"
+  VERSION = "0.7.45"
 end

data/lib/pdfrb.rb

@@ -58,6 +58,20 @@ module Pdfrb
   autoload :CLI, "pdfrb/cli"
 
   require "logger"
+  require "stringio"
+
+  class << self
+    # Open a PDF file (path or block over the file handle).
+    def open(path, **, &)
+      Document.open(path, **, &)
+    end
+
+    # Parse PDF bytes/IO into a Document.
+    def parse(io, **)
+      io = StringIO.new(io.b) if io.is_a?(String)
+      Document.new(io: io, **)
+    end
+  end
 
   def self.logger
     @logger ||= Logger.new($stderr, level: Logger::WARN)

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pdfrb
 version: !ruby/object:Gem::Version
-  version: 0.7.43
+  version: 0.7.45
 platform: ruby
 authors:
 - Ribose Inc.