Files
hakorune/crates/nyrt/src/plugin/map.rs

185 lines
6.9 KiB
Rust
Raw Normal View History

2025-09-11 11:19:55 +09:00
// --- AOT ObjectModule dotted-name exports (Map) ---
// Provide dotted symbol names expected by ObjectBuilder lowering for MapBox operations.
// size: (handle) -> i64
#[export_name = "nyash.map.size_h"]
pub extern "C" fn nyash_map_size_h(handle: i64) -> i64 {
use nyash_rust::jit::rt::handles;
if std::env::var("NYASH_LLVM_MAP_DEBUG").ok().as_deref() == Some("1") {
eprintln!("[MAP] size_h(handle={})", handle);
}
2025-09-11 11:19:55 +09:00
if handle <= 0 {
return 0;
}
if let Some(obj) = handles::get(handle as u64) {
if let Some(map) = obj
.as_any()
.downcast_ref::<nyash_rust::boxes::map_box::MapBox>()
{
if let Some(ib) = map
.size()
.as_any()
.downcast_ref::<nyash_rust::box_trait::IntegerBox>()
{
if std::env::var("NYASH_LLVM_MAP_DEBUG").ok().as_deref() == Some("1") {
eprintln!("[MAP] size_h => {}", ib.value);
}
2025-09-11 11:19:55 +09:00
return ib.value;
}
}
}
0
}
// get_h: (map_handle, key_i64) -> value_handle
#[export_name = "nyash.map.get_h"]
pub extern "C" fn nyash_map_get_h(handle: i64, key: i64) -> i64 {
use nyash_rust::{
box_trait::{IntegerBox, NyashBox},
jit::rt::handles,
};
if std::env::var("NYASH_LLVM_MAP_DEBUG").ok().as_deref() == Some("1") {
eprintln!("[MAP] get_h(handle={}, key={})", handle, key);
}
2025-09-11 11:19:55 +09:00
if handle <= 0 {
return 0;
}
if let Some(obj) = handles::get(handle as u64) {
if let Some(map) = obj
.as_any()
.downcast_ref::<nyash_rust::boxes::map_box::MapBox>()
{
let kbox: Box<dyn NyashBox> = Box::new(IntegerBox::new(key));
let v = map.get(kbox);
let arc: std::sync::Arc<dyn NyashBox> = std::sync::Arc::from(v);
let h = handles::to_handle(arc);
if std::env::var("NYASH_LLVM_MAP_DEBUG").ok().as_deref() == Some("1") {
eprintln!("[MAP] get_h => handle {}", h);
}
2025-09-11 11:19:55 +09:00
return h as i64;
}
}
0
}
// get_hh: (map_handle, key_handle) -> value_handle
#[export_name = "nyash.map.get_hh"]
🔍 Research: GPT-5-Codex capabilities and GitHub PR integration ## Summary Investigated OpenAI's new GPT-5-Codex model and Codex GitHub PR review integration capabilities. ## GPT-5-Codex Analysis ### Benchmark Performance (Good) - SWE-bench Verified: 74.5% (vs GPT-5's 72.8%) - Refactoring tasks: 51.3% (vs GPT-5's 33.9%) - Code review: Higher developer ratings ### Real-World Issues (Concerning) - Users report degraded coding performance - Scripts that previously worked now fail - Less consistent than GPT-4.5 - Longer response times (minutes vs instant) - "Creatively and emotionally flat" - Basic errors (e.g., counting letters incorrectly) ### Key Finding Classic case of "optimizing for benchmarks vs real usability" - scores well on tests but performs poorly in practice. ## Codex GitHub PR Integration ### Setup Process 1. Enable MFA and connect GitHub account 2. Authorize Codex GitHub app for repos 3. Enable "Code review" in repository settings ### Usage Methods - **Manual**: Comment '@codex review' in PR - **Automatic**: Triggers when PR moves from draft to ready ### Current Limitations - One-way communication (doesn't respond to review comments) - Prefers creating new PRs over updating existing ones - Better for single-pass reviews than iterative feedback ## 'codex resume' Feature New session management capability: - Resume previous codex exec sessions - Useful for continuing long tasks across days - Maintains context from interrupted work 🐱 The investigation reveals that while GPT-5-Codex shows benchmark improvements, practical developer experience has declined - a reminder that metrics don't always reflect real-world utility\!
2025-09-16 16:28:25 +09:00
pub extern "C" fn nyash_map_get_hh(handle: i64, key_any: i64) -> i64 {
use nyash_rust::{box_trait::{NyashBox, IntegerBox}, jit::rt::handles};
if handle <= 0 { return 0; }
if let Some(obj) = handles::get(handle as u64) {
if let Some(map) = obj.as_any().downcast_ref::<nyash_rust::boxes::map_box::MapBox>() {
let key_box: Box<dyn NyashBox> = if key_any > 0 {
if let Some(k) = handles::get(key_any as u64) { k.clone_box() } else { Box::new(IntegerBox::new(key_any)) }
} else { Box::new(IntegerBox::new(key_any)) };
let v = map.get(key_box);
2025-09-11 11:19:55 +09:00
let arc: std::sync::Arc<dyn NyashBox> = std::sync::Arc::from(v);
let h = handles::to_handle(arc);
return h as i64;
}
}
0
}
🔍 Research: GPT-5-Codex capabilities and GitHub PR integration ## Summary Investigated OpenAI's new GPT-5-Codex model and Codex GitHub PR review integration capabilities. ## GPT-5-Codex Analysis ### Benchmark Performance (Good) - SWE-bench Verified: 74.5% (vs GPT-5's 72.8%) - Refactoring tasks: 51.3% (vs GPT-5's 33.9%) - Code review: Higher developer ratings ### Real-World Issues (Concerning) - Users report degraded coding performance - Scripts that previously worked now fail - Less consistent than GPT-4.5 - Longer response times (minutes vs instant) - "Creatively and emotionally flat" - Basic errors (e.g., counting letters incorrectly) ### Key Finding Classic case of "optimizing for benchmarks vs real usability" - scores well on tests but performs poorly in practice. ## Codex GitHub PR Integration ### Setup Process 1. Enable MFA and connect GitHub account 2. Authorize Codex GitHub app for repos 3. Enable "Code review" in repository settings ### Usage Methods - **Manual**: Comment '@codex review' in PR - **Automatic**: Triggers when PR moves from draft to ready ### Current Limitations - One-way communication (doesn't respond to review comments) - Prefers creating new PRs over updating existing ones - Better for single-pass reviews than iterative feedback ## 'codex resume' Feature New session management capability: - Resume previous codex exec sessions - Useful for continuing long tasks across days - Maintains context from interrupted work 🐱 The investigation reveals that while GPT-5-Codex shows benchmark improvements, practical developer experience has declined - a reminder that metrics don't always reflect real-world utility\!
2025-09-16 16:28:25 +09:00
2025-09-11 11:19:55 +09:00
// set_h: (map_handle, key_i64, val) -> i64 (ignored/0)
#[export_name = "nyash.map.set_h"]
pub extern "C" fn nyash_map_set_h(handle: i64, key: i64, val: i64) -> i64 {
use nyash_rust::{
box_trait::{IntegerBox, NyashBox},
jit::rt::handles,
};
if std::env::var("NYASH_LLVM_MAP_DEBUG").ok().as_deref() == Some("1") {
eprintln!("[MAP] set_h(handle={}, key={}, val={})", handle, key, val);
}
2025-09-11 11:19:55 +09:00
if handle <= 0 {
return 0;
}
if let Some(obj) = handles::get(handle as u64) {
if let Some(map) = obj
.as_any()
.downcast_ref::<nyash_rust::boxes::map_box::MapBox>()
{
let kbox: Box<dyn NyashBox> = Box::new(IntegerBox::new(key));
let vbox: Box<dyn NyashBox> = if val > 0 {
if let Some(o) = handles::get(val as u64) {
o.clone_box()
} else {
Box::new(IntegerBox::new(val))
}
} else {
Box::new(IntegerBox::new(val))
};
let _ = map.set(kbox, vbox);
if std::env::var("NYASH_LLVM_MAP_DEBUG").ok().as_deref() == Some("1") {
let sz = map
.size()
.as_any()
.downcast_ref::<nyash_rust::box_trait::IntegerBox>()
.map(|i| i.value)
.unwrap_or(-1);
eprintln!("[MAP] set_h done; size now {}", sz);
}
2025-09-11 11:19:55 +09:00
return 0;
}
}
0
}
🔍 Research: GPT-5-Codex capabilities and GitHub PR integration ## Summary Investigated OpenAI's new GPT-5-Codex model and Codex GitHub PR review integration capabilities. ## GPT-5-Codex Analysis ### Benchmark Performance (Good) - SWE-bench Verified: 74.5% (vs GPT-5's 72.8%) - Refactoring tasks: 51.3% (vs GPT-5's 33.9%) - Code review: Higher developer ratings ### Real-World Issues (Concerning) - Users report degraded coding performance - Scripts that previously worked now fail - Less consistent than GPT-4.5 - Longer response times (minutes vs instant) - "Creatively and emotionally flat" - Basic errors (e.g., counting letters incorrectly) ### Key Finding Classic case of "optimizing for benchmarks vs real usability" - scores well on tests but performs poorly in practice. ## Codex GitHub PR Integration ### Setup Process 1. Enable MFA and connect GitHub account 2. Authorize Codex GitHub app for repos 3. Enable "Code review" in repository settings ### Usage Methods - **Manual**: Comment '@codex review' in PR - **Automatic**: Triggers when PR moves from draft to ready ### Current Limitations - One-way communication (doesn't respond to review comments) - Prefers creating new PRs over updating existing ones - Better for single-pass reviews than iterative feedback ## 'codex resume' Feature New session management capability: - Resume previous codex exec sessions - Useful for continuing long tasks across days - Maintains context from interrupted work 🐱 The investigation reveals that while GPT-5-Codex shows benchmark improvements, practical developer experience has declined - a reminder that metrics don't always reflect real-world utility\!
2025-09-16 16:28:25 +09:00
// set_hh: (map_handle, key_any: handle or i64, val_any: handle or i64) -> i64
#[export_name = "nyash.map.set_hh"]
pub extern "C" fn nyash_map_set_hh(handle: i64, key_any: i64, val_any: i64) -> i64 {
use nyash_rust::{box_trait::{NyashBox, IntegerBox}, jit::rt::handles};
if handle <= 0 { return 0; }
if let Some(obj) = handles::get(handle as u64) {
if let Some(map) = obj.as_any().downcast_ref::<nyash_rust::boxes::map_box::MapBox>() {
let kbox: Box<dyn NyashBox> = if key_any > 0 {
if let Some(k) = handles::get(key_any as u64) { k.clone_box() } else { Box::new(IntegerBox::new(key_any)) }
} else { Box::new(IntegerBox::new(key_any)) };
let vbox: Box<dyn NyashBox> = if val_any > 0 {
if let Some(v) = handles::get(val_any as u64) { v.clone_box() } else { Box::new(IntegerBox::new(val_any)) }
} else { Box::new(IntegerBox::new(val_any)) };
let _ = map.set(kbox, vbox);
return 0;
}
}
0
}
// has_hh: (map_handle, key_any: handle or i64) -> i64 (0/1)
#[export_name = "nyash.map.has_hh"]
pub extern "C" fn nyash_map_has_hh(handle: i64, key_any: i64) -> i64 {
use nyash_rust::{box_trait::{NyashBox, IntegerBox, BoolBox}, jit::rt::handles};
if handle <= 0 { return 0; }
if let Some(obj) = handles::get(handle as u64) {
if let Some(map) = obj.as_any().downcast_ref::<nyash_rust::boxes::map_box::MapBox>() {
let kbox: Box<dyn NyashBox> = if key_any > 0 {
if let Some(k) = handles::get(key_any as u64) { k.clone_box() } else { Box::new(IntegerBox::new(key_any)) }
} else { Box::new(IntegerBox::new(key_any)) };
let v = map.has(kbox);
if let Some(b) = v.as_any().downcast_ref::<BoolBox>() { return if b.value { 1 } else { 0 }; }
}
}
0
}
2025-09-11 11:19:55 +09:00
// has_h: (map_handle, key_i64) -> i64 (0/1)
#[export_name = "nyash.map.has_h"]
pub extern "C" fn nyash_map_has_h(handle: i64, key: i64) -> i64 {
use nyash_rust::{box_trait::IntegerBox, jit::rt::handles};
if handle <= 0 {
return 0;
}
if let Some(obj) = handles::get(handle as u64) {
if let Some(map) = obj
.as_any()
.downcast_ref::<nyash_rust::boxes::map_box::MapBox>()
{
let kbox = Box::new(IntegerBox::new(key));
🔍 Research: GPT-5-Codex capabilities and GitHub PR integration ## Summary Investigated OpenAI's new GPT-5-Codex model and Codex GitHub PR review integration capabilities. ## GPT-5-Codex Analysis ### Benchmark Performance (Good) - SWE-bench Verified: 74.5% (vs GPT-5's 72.8%) - Refactoring tasks: 51.3% (vs GPT-5's 33.9%) - Code review: Higher developer ratings ### Real-World Issues (Concerning) - Users report degraded coding performance - Scripts that previously worked now fail - Less consistent than GPT-4.5 - Longer response times (minutes vs instant) - "Creatively and emotionally flat" - Basic errors (e.g., counting letters incorrectly) ### Key Finding Classic case of "optimizing for benchmarks vs real usability" - scores well on tests but performs poorly in practice. ## Codex GitHub PR Integration ### Setup Process 1. Enable MFA and connect GitHub account 2. Authorize Codex GitHub app for repos 3. Enable "Code review" in repository settings ### Usage Methods - **Manual**: Comment '@codex review' in PR - **Automatic**: Triggers when PR moves from draft to ready ### Current Limitations - One-way communication (doesn't respond to review comments) - Prefers creating new PRs over updating existing ones - Better for single-pass reviews than iterative feedback ## 'codex resume' Feature New session management capability: - Resume previous codex exec sessions - Useful for continuing long tasks across days - Maintains context from interrupted work 🐱 The investigation reveals that while GPT-5-Codex shows benchmark improvements, practical developer experience has declined - a reminder that metrics don't always reflect real-world utility\!
2025-09-16 16:28:25 +09:00
let v = map.has(kbox);
if let Some(b) = v.as_any().downcast_ref::<nyash_rust::box_trait::BoolBox>() { return if b.value { 1 } else { 0 }; }
2025-09-11 11:19:55 +09:00
}
}
0
}