message_segment_calculator 1.3.0

message_segment_calculator: ^1.3.0 copied to clipboard

Brings the package in line with Twilio's official calculator (TwilioDevEd/message-segment-calculator v1.3.0). Verified by running both calculators on more than 10,000 messages (emoji, flags, skin tones, ZWJ sequences, accents, CJK, GSM-7 extension characters, line breaks, smart encoding, forced encodings) and comparing every output; they agree on all of them.

Fixed #

• Grapheme splitting now matches Twilio — Replaced package:characters with a Dart port of grapheme-splitter 1.0.4, the splitter Twilio uses. Newer emoji sequences (e.g. 🏴‍☠️, 🫱🏼‍🫲🏿, 🐦‍🔥) are split the same way, so the character and segment counts are the same as Twilio's.
• Spaces and line breaks were reported as non-GSM — EncodedChar.isGSM7 trimmed the character before checking it, so ' ', '\n' and '\r' were false and showed up in getNonGsmCharacters().
• GSM extension characters in UCS-2 messages — |, ^, {, €, … now keep their single UTF-16 code unit in codeUnits when the message is UCS-2, instead of the 2-unit GSM-7 escape sequence (Twilio #57/#61).
• Line break style — A message with only CRLF line breaks is now reported as LineBreakStyle.crlf instead of lfCrlf (Twilio #67).
• Smart encoding — ‹ and › now map to < and > (they were swapped, Twilio #58), and ZERO WIDTH NON-JOINER (U+200C) is removed. Smart encoding now iterates Unicode code points, like Twilio.
• Line break warning now uses Twilio's exact wording.
• Segment.addHeader() sets hasUserDataHeader to false like Twilio (hasTwilioReservedBits is true). Segment counts are unaffected.

Added #

• RcsSegmentedMessage — RCS billing (UTF-8 bytes): US messages are split into 160-byte "Rich" segments; international messages are "Basic" (≤160 bytes) or "Single" (>160 bytes, capacity 1600).
• countUtf8Bytes() — UTF-8 byte length of a message.
• Port of Twilio's full test suite (test/twilio_parity_test.dart).

Changed #

• The package no longer depends on package:characters.

Added #

• remainingCharsInSegment property — Returns how many characters can still be added to the current (last) segment before a new segment is created. Useful for building real-time character counters and SMS cost previews.
• maxCharsPerSegment property — Returns the maximum number of characters that fit in a single segment for the current encoding. Accounts for the User Data Header overhead in multi-segment messages (GSM-7: 160→153, UCS-2: 70→67).
• 8 new unit tests — Comprehensive tests covering empty messages, exact boundary, single/multi-segment transitions, and extended GSM-7 characters for both new properties.

Fixed #

• CI/CD: skip example directory during publish — The example/ folder is a Flutter project which caused dart pub get to fail with exit code 69. CI now uses --no-example for dependency resolution and scopes format/analyze to lib/ and test/ only.

Fixed #

• Replaced flutter_lints with lints — The flutter_lints dev dependency pulled in the Flutter SDK, causing dart pub publish to fail on CI with exit code 69. Replaced with the pure Dart lints package.

Breaking Changes #

• Converted to a pure Dart package — Removed Flutter SDK dependency (flutter, flutter_test), making the package usable in any Dart project (CLI, server-side, Flutter, etc.).
• Renamed SmsEncoding enum to SmsEncodingMode — The previous SmsEncoding enum (with gsm7, ucs2, auto) is now SmsEncodingMode. A new SmsEncoding enum (with gsm7, ucs2 only) represents the resolved encoding.
• Removed ValidEncodingValues enum — Replaced by the new SmsEncoding enum.
• Replaced string-based encoding with enums — EncodedChar now uses SmsEncoding enum instead of raw strings like 'gsm7' / 'ucs2'.

Added #

• Introduced SegmentElement sealed class — A new base type for EncodedChar and UserDataHeader, enabling type-safe segment composition and proper polymorphism.
• Added new test cases — Tests for long GSM-7 multi-segment messages and long UCS-2 multi-segment messages.
• Added toString() override on Segment — Useful for debugging segment details.
• Added repository, issue_tracker, and topics metadata to pubspec.yaml for better pub.flutter-io.cn discoverability.

Changed #

• Replaced grapheme_splitter dependency with characters — Uses the official Dart characters package for grapheme splitting, reducing dependency bloat.
• Replaced flutter_test with test package — Tests now use the standard Dart test package.
• Type-safe segment elements — Segment._elements changed from List<dynamic> to List<SegmentElement>, add() and removeLast() now use SegmentElement type.
• Simplified line break processing — Grapheme line break handling now uses expand() instead of fold().
• Simplified _hasAnyUCSCharacters — Refactored from a verbose loop to a concise any() expression.
• Improved sizeInBits() calculations — Segment.sizeInBits() now correctly sums all element sizes (previously ignored UserDataHeader sizes in computation).
• Fixed addHeader() bug — hasUserDataHeader is now correctly set to true (was previously set to false by mistake).
• Fixed typo in filename — Renamed enchoded_char.dart to encoded_char.dart.
• Code cleanup — Replaced var/explicit types with final where appropriate, used expression body syntax, and improved formatting throughout.

Removed #

• Removed grapheme_splitter dependency — Replaced by the characters package.
• Removed Flutter SDK dependency — Package is now pure Dart.

Changed #

• fix: importing issue fixed

Changed #

• fix : removed unused functions
• fix : removed unused variables
• fix : removed unused imports
• fix : Code format and refractor

Changed #

• fix : segment calculation logics
• improved usability
• fix : encoding logic
• Updated the README with detailed descriptions and usage examples.

Changed #

• Code format and refractor

Added #

• Added the SegmentedMessage class to handle the segmentation logic for dividing a message into multiple SMS segments.
• Added the EncodedChar class to represent individual characters and their encoding properties.
• Added the Segment class to manage individual segments, handle additions and removals, and calculate segment sizes.
• Added the UserDataHeader class to represent the User Data Header required for concatenated SMS messages.
• Enabled public_member_api_docs lint to ensure all public API members are documented.
• Introduced example usage in README for easy integration.

Changed #

• Improved error handling in the main entry point for better debugging and usability.
• Updated the README with detailed descriptions and usage examples.

• TODO: Describe initial release.