dearwolf/aider
1
0
Fork 0
mirror of https://github.com/Aider-AI/aider.git synced 2025-05-25 14:55:00 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

precise edit blocks prompting language

Browse source
This commit is contained in:
Paul Gauthier 2023-08-09 11:57:00 -03:00
parent 2c53c153b8
commit 9e64e7bb9d
2 changed files with 6 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

5
aider/coders/editblock_prompts.py
View file

@ -32,10 +32,15 @@ some/dir/example.py
>>>>>>> UPDATED >>>>>>> UPDATED
{fence[1]} {fence[1]}
A program will parse the edit blocks you generate and replace the ORIGINAL lines with the UPDATED lines.
So edit blocks must be precise and unambiguous!
Every *edit block* must be fenced with {fence[0]}...{fence[1]} with the correct code language. Every *edit block* must be fenced with {fence[0]}...{fence[1]} with the correct code language.
Every *edit block* must start with the full path! *NEVER* propose edit blocks for *read-only* files. Every *edit block* must start with the full path! *NEVER* propose edit blocks for *read-only* files.
The ORIGINAL section must be an *exact set of sequential lines* from the file!
NEVER SKIP LINES in the ORIGINAL section! NEVER SKIP LINES in the ORIGINAL section!
NEVER ELIDE LINES AND REPLACE THEM WITH A COMMENT!
NEVER OMIT ANY WHITESPACE in the ORIGINAL section! NEVER OMIT ANY WHITESPACE in the ORIGINAL section!
Edits to different parts of a file each need their own *edit block*. Edits to different parts of a file each need their own *edit block*.

5
aider/sendchat.py
View file

@ -13,7 +13,7 @@ from openai.error import (
Timeout, Timeout,
) )
CACHE_PATH = ".aider.send.cache.v1" CACHE_PATH = "~/.aider.send.cache.v1"
CACHE = Cache(CACHE_PATH) CACHE = Cache(CACHE_PATH)
@ -54,11 +54,8 @@ def send_with_retries(model, messages, functions, stream):
hash_object = hashlib.sha1(key) hash_object = hashlib.sha1(key)
if not stream and key in CACHE: if not stream and key in CACHE:
print("hit", key)
return hash_object, CACHE[key] return hash_object, CACHE[key]
print("miss", key)
res = openai.ChatCompletion.create(**kwargs) res = openai.ChatCompletion.create(**kwargs)
if not stream: if not stream: