fosster/libb20
1
0
Fork 0
Code Issues Pull requests Projects Releases 1 Packages Wiki Activity Actions

feat: libb20 C90 #11

Merged
sezieru merged 10 commits from feat/lib/c90/main into main 2025-11-03 12:31:59 +00:00
Conversation 0 Commits 10 Files changed 7 +1229 -865
sezieru commented 2025-11-03 12:31:43 +00:00
Owner
Copy link

🧮 libb20 – C90 Library Implementation & Reference CLI

This PR marks the formal transition of libb20 from a proof-of-concept encoder to a proper reusable C90 library, establishing the structure that all future language ports will follow.

🧱 What’s New

🧩 Core Library (libb20.h / libb20.c)

  • Full ANSI/ISO-C90 implementation of the base-20 codec.
  • Shared internal helpers for UTF-8, base-256 math, and block encoding.
  • Clean public interface with explicit ownership and error model.
  • Added b20_status_str() for human-readable diagnostics.
  • All operations are big-endian (MSB-first) and re-entrant.

🔧 Utility Layer (libb20_util.h / .c)

  • Optional module for UTF-8 parsing and encoding convenience.
  • Provides functions to safely interpret CLI input into code points.
  • Keeps the main library Unicode-agnostic and minimal.

🖥️ Reference CLI (kb20-c90ref)

  • New command-line tool built entirely on the public API.

    • encode / decode
    • --text or --bin modes
    • --zero <code point> or --alphabet <20 code points>

  • Replaces the previous standalone reference implementation (kb20c).

  • Reads from stdin, writes to stdout, returns clear status codes.

🧰 Build System

  • Updated Makefile to compile and link the library and CLI cleanly.
  • Added make clean target.
  • Retired the legacy reference implementation.

🧪 Verification

  • Cross-checked output against the original reference:

    • Text encoding produces identical results for all test inputs.
    • Binary-digits mode round-trips byte-for-byte.

  • Verified correct decoding for variable-length tails.

  • Confirmed proper alphabet validation and duplicate detection.

  • Verified all error codes via malformed inputs.

🧭 Future Work

  • Integrate lightweight tests for each error path and block length edge case.
  • Begin formal documentation of the binary format and byte layout.
  • Start work on the C++, Python, and TypeScript bindings.

📜 Summary

This PR graduates libb20 into its first stable architectural form:
a clean, standards-compliant base-20 codec library designed for reuse,
teaching, and interoperability across languages.

# 🧮 libb20 – C90 Library Implementation & Reference CLI This PR marks the formal transition of **libb20** from a proof-of-concept encoder to a **proper reusable C90 library**, establishing the structure that all future language ports will follow. --- ## 🧱 What’s New ### 🧩 Core Library (`libb20.h` / `libb20.c`) - Full ANSI/ISO-C90 implementation of the base-20 codec. - Shared internal helpers for UTF-8, base-256 math, and block encoding. - Clean public interface with explicit ownership and error model. - Added `b20_status_str()` for human-readable diagnostics. - All operations are **big-endian (MSB-first)** and re-entrant. ### 🔧 Utility Layer (`libb20_util.h` / `.c`) - Optional module for UTF-8 parsing and encoding convenience. - Provides functions to safely interpret CLI input into code points. - Keeps the main library Unicode-agnostic and minimal. ### 🖥️ Reference CLI (`kb20-c90ref`) - New command-line tool built entirely on the public API. - `encode` / `decode` - `--text` or `--bin` modes - `--zero <code point>` or `--alphabet <20 code points>` - Replaces the previous standalone reference implementation (`kb20c`). - Reads from **stdin**, writes to **stdout**, returns clear status codes. ### 🧰 Build System - Updated `Makefile` to compile and link the library and CLI cleanly. - Added `make clean` target. - Retired the legacy reference implementation. --- ## 🧪 Verification - Cross-checked output against the original reference: - Text encoding produces identical results for all test inputs. - Binary-digits mode round-trips byte-for-byte. - Verified correct decoding for variable-length tails. - Confirmed proper alphabet validation and duplicate detection. - Verified all error codes via malformed inputs. --- ## 🧭 Future Work - Integrate lightweight tests for each error path and block length edge case. - Begin formal documentation of the binary format and byte layout. - Start work on the **C++**, **Python**, and **TypeScript** bindings. --- ## 📜 Summary This PR graduates libb20 into its **first stable architectural form**: a clean, standards-compliant base-20 codec library designed for reuse, teaching, and interoperability across languages.
sezieru self-assigned this 2025-11-03 12:31:43 +00:00 
- Introduce b20_* interfaces for test and binary-digits modes
- Alphabet config via ZERO cp or 20-scalar map
- Explicit error mode (b20_status) and allocator hook (b20_free)
- Implement text and binary-digits codecs using big-endian folds
- Well-factored static helpers: UTF-8 (internal), big-int ops, block enc/dec, len<->digits
- C90/ANSI compliant, reentrant, no globals
- Subcommands encode/decode; --text (default) or --bin
- Text mode supports --zero <codepoint> or --alphabet <20 codepoints>
- Streams stdin -> stdout; human-readable errors
- Map b20_status to stable strings for UIs/CLIs/logging
- b20_utf8_decode_one/encode_one, b20_utf8_parse_one_exact, b20_utf8_parse_20
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
sezieru
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference: fosster/libb20#11
No description provided.
Delete branch "feat/lib/c90/main"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?