tans-parser 0.1.3 → 0.1.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: fc3b07fbdd59e595af3d07d3091cdb75b9faa0c4daeaaaf3232f7787d35406e4
-  data.tar.gz: 75f354d01a6881cc8bb702bb900b41b2f654ecae8b7d5033bc439ad876eeb97a
+  metadata.gz: 907b90ef203876bd99cc0dbca8eb6b184ed6472b580c6d2918e31469d0f7cc12
+  data.tar.gz: b828065265563752bb2acf5ef380c2ba805f4b4e785fc602337c2b5048267824
 SHA512:
-  metadata.gz: 7d79356c65433991718c0c1854faa58f5a077a906ec653f13202d241b37fc01bb3fe253de8ba4fa9deb3b7bdee17965a1b2d2476a3b782c0a849c8b0696c4a65
-  data.tar.gz: 05045d158716d7e4ef9f87ecd2d4936735d1de681fb17cb97835a078b31654c9d0ed47022b610db7824d2a7408e82b66c34df738f1bc3b0ad7e67e40bfb4e111
+  metadata.gz: ae5ea0f42c3663d0edfc35e86bb11064396099f73beebcdcf2fe5d43cfc2b65d810ab2dc055cf5d186a3571aa11703abc006962407a30b979c1a7d4d6e46f3f7
+  data.tar.gz: b3379404c4cee09f3e49ceeaaa0e3ec90322a665fbb671a7d5e19506759b72e338fce4e782cc2017779d7812eadfddc098df82ba6bd0cb4ecec50a283d4ce044

data/CHANGELOG.md

@@ -1,14 +1,27 @@
 # CHANGELOG
 
+## 0.1.4
+
+- **Unicode width support** — correct display width for CJK, emoji, and combining characters:
+  - `unicode-display_width` gem as runtime dependency (~> 2.5)
+  - `:width` key in cell hash (1 or 2) and `default_cell`
+  - Cursor advances by display width instead of always +1
+  - Wide chars (CJK/emoji) clear continuation cells
+  - Combining characters (zero-width) appended to previous cell
+  - Bugfix: parse loop uses `bytesize` to handle multi-byte chars at start of string
+- 4 new tests, 329 total, 100% line and branch coverage maintained
+
 ## 0.1.3
 
-- **Dialog recognition** — added support for rounded corners (`╭╮╰╯`) and double-line (`╔╗╚╝`) box-drawing characters
+- **Dialog recognition** — extended box-drawing character support and titled borders:
+  - Added rounded corners (`╭╮╰╯`) and double-line (`╔╗╚╝`)
   - `TOP_LEFT_CORNERS` extended with `╭╔╓╒`
   - `dialog_top_width` extended with `╮╗╖╕` (top-right) and `═` (double horizontal)
+  - Supports titled borders: finds first top-right corner anywhere on line (e.g. `╭─ Commands ─╮`)
 - **Statusbar recognition** — more flexible detection:
-  - Now checks last 2 rows instead of only the last row
+  - Checks last 2 rows instead of only the last row
   - Fallback: detects last row as statusbar if it has ≥30 characters of content, even without background color info
-  - Handles Karat-style footers (`? for shortcuts | mock ctx ░░░░░░░░░░ 0%`)
+  - Separator-preceded footers: scans all rows for footer after `───` separator line (Karat-style)
 - **Custom role registration** — `State#annotate_role(role, row:, col:, width:, height:, text:, **extra)`:
   - Manually annotate grid regions with semantic roles
   - `Selector#detect_annotations` picks them up during `scan` alongside auto-detected elements
@@ -16,9 +29,10 @@
 - **State#diff** — cell-level comparison between two State instances:
   - `diff(other_state)` — compares all 7 cell keys (`char`, `fg`, `bg`, `bold`, `italic`, `underline`, `blink`)
   - `diff(other_state, chars_only: true)` — compares only `:char`, ignores style/color changes
+  - `diff(other_state, ignore_rows: [2, 5])` — skips specified rows (e.g. cursor/prompt lines)
   - Handles different grid sizes (fills missing cells with `DEFAULT_CELL`)
   - Accepts raw hash or State object
-- 19 new tests, 319 total, 100% line and branch coverage maintained
+- 25 new tests, 325 total, 100% line and branch coverage maintained
 
 ## 0.1.2
 

data/README.md

@@ -173,6 +173,9 @@ diff = before.diff(after)
 diff = before.diff(after, chars_only: true)
 # Only reports actual character differences
 
+# Ignore specific rows — useful for cursor/prompt lines
+diff = before.diff(after, chars_only: true, ignore_rows: [prompt_row])
+
 # Accepts raw hash as argument
 diff = before.diff({size: {rows: 5, cols: 10}, cursor: {...}, rows: [...]})
 ```

data/lib/tans_parser/ansi_parser.rb

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+require "unicode/display_width"
+
 module TansParser
   # Parses raw terminal output (ANSI escape sequences + text) into a
   # structured state representation.
@@ -154,11 +156,11 @@ module TansParser
       processed = raw
 
       i = 0
-      while i < processed.length
+      while i < processed.bytesize
         if processed[i] == "\e" && processed[i + 1] == "["
           # Find end of CSI sequence
           j = i + 2
-          j += 1 while j < processed.length && !processed[j].match?(/[A-HJ-KP-SX@`fhlmnRrsuq]/)
+          j += 1 while j < processed.bytesize && !processed[j].match?(/[A-HJ-KP-SX@`fhlmnRrsuq]/)
           seq = processed[i..j]
 
           dsr, new_saved, action = _apply_csi(seq, cursor, attrs, grid, rows, cols, saved_cursor, scroll_region)
@@ -279,14 +281,29 @@ module TansParser
         elsif (char, char_len = _utf8_char_at(processed, i))
           # Printable character (including multi-byte UTF-8)
           # cursor row/col are always clamped within bounds
-          cell = grid[cursor[:row]][cursor[:col]]
           current_charset = (active_charset == :g1 ? g1_charset : g0_charset)
           mapped_char = char
           mapped_char = DEC_MAP[char] if current_charset == :dec && DEC_MAP.key?(char)
-          cell[:char] = mapped_char
-          cell.merge!(attrs)
-          cursor[:col] += 1
-          cursor[:col] = cols - 1 if cursor[:col] >= cols
+
+          char_width = Unicode::DisplayWidth.of(mapped_char)
+          if char_width.zero? && cursor[:col].positive?
+            # Combining character — append to previous cell
+            prev_cell = grid[cursor[:row]][cursor[:col] - 1]
+            prev_cell[:char] = prev_cell[:char] + mapped_char
+          else
+            cell = grid[cursor[:row]][cursor[:col]]
+            cell[:char] = mapped_char
+            cell[:width] = char_width
+            cell.merge!(attrs)
+            # Clear continuation cells for wide characters
+            (1...char_width).each do |off|
+              cont = grid[cursor[:row]][cursor[:col] + off]
+              cont[:char] = ""
+              cont[:width] = 0
+            end
+            cursor[:col] += char_width
+            cursor[:col] = cols - 1 if cursor[:col] >= cols
+          end
           i += char_len
         else # rubocop:disable Lint/DuplicateBranch
           i += 1
@@ -691,7 +708,7 @@ module TansParser
     # rubocop:enable Metrics/CyclomaticComplexity
 
     def self.default_cell
-      { char: " ", fg: "default", bg: "default", bold: false, italic: false, underline: false, blink: false }
+      { char: " ", fg: "default", bg: "default", bold: false, italic: false, underline: false, blink: false, width: 1 }
     end
 
     # Extract a single UTF-8 character at position i in a binary string.

data/lib/tans_parser/state.rb

@@ -124,7 +124,7 @@ module TansParser
     end
 
     DEFAULT_CELL = { char: " ", fg: "default", bg: "default",
-                     bold: false, italic: false, underline: false, blink: false, }.freeze
+                     bold: false, italic: false, underline: false, blink: false, width: 1, }.freeze
 
     # Compare this state with another State and return cell-level differences.
     # With chars_only: true, only differences in the :char key are reported.

data/lib/tans_parser/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module TansParser
-  VERSION = "0.1.3"
+  VERSION = "0.1.4"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: tans-parser
 version: !ruby/object:Gem::Version
-  version: 0.1.3
+  version: 0.1.4
 platform: ruby
 authors:
 - Haluk Durmus
@@ -121,6 +121,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '0.22'
+- !ruby/object:Gem::Dependency
+  name: unicode-display_width
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.5'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.5'
 description: tans-parser parses raw terminal output with ANSI escape sequences into
   a structured grid representation with per-cell attributes (char, fg, bg, bold, italic,
   underline, blink). Includes a query API (State) for text search, color inspection,