[elpa] externals/llm updated (efe218ac13 -> 478afbcb41)

[ELPA Syncer]

elpasync pushed a change to branch externals/llm.

      from  efe218ac13 Set version to 0.14.2
       new  f33475eeae Add plz and media type and event source extensions
       new  a9cd296cd8 Add llm-request-plz.el
       new  1f3b018dcb Merge pull request #26 from r0man/plz
       new  829bedabe6 Support for vertex embeddings, adding callback 
compatibility
       new  41d0889bcb Make sync use of plz return data instead of object, 
converted Gemini
       new  2ac956a060 Add support for the application/x-ndjson media type
       new  84678edfae Merge pull request #28 from r0man/plz
       new  9541d34656 Enable streaming in Open AI with plz modifications
       new  c9ab8664ce Support function streaming with Open AI & plz
       new  63f2b8ffbc Merge branch 'main' into plz
       new  3988fecb53 Make separate function for event streaming, w/ 
client-side handlers
       new  7f73a59fd1 Remove media-types from llm-request-plz-async
       new  9ce3d9e003 Port claude to llm-request-plz
       new  f9213b981c Use the plz request module for everything in Claude
       new  f41c55a44e Switch Open AI completely to plz, fix async request 
method
       new  83f05280fc Strip plz changes and add JSON array stream media type
       new  5bb3dadf25 Handle empty choices
       new  cdbb41528c Fix issue advancing the process buffer
       new  3441784ae2 Fix error handling.
       new  6f9c604e58 Decode body and chunks using a coding system
       new  62495de57f Don't decode error response twice
       new  2a0651adc2 Merge pull request #29 from r0man/plz
       new  03afb1d795 Use symbols for event types & improve timeout handling
       new  e07c84b81b Improve Curl/HTTP error handling
       new  2ea07f9197 Merge pull request #30 from r0man/plz-event-type-symbol
       new  069bd193ef Adding Gemini & Vertex streaming using plz
       new  27884dcb32 Port ollama to plz
       new  99814b533d Changes to get at least some of Gemini / Vertex working
       new  623bceda67 Fix issue with JSON array parser nor emitting all objects
       new  7919563997 Merge pull request #31 from r0man/plz
       new  40909718cc Fix function calling for Gemini & Vertex, and streaming 
for Gemini
       new  385ce9cdac Use the streaming endpoint for the Vertex provider
       new  9c07e22659 Kill the response buffer when the request completed
       new  a61eb88c53 Merge pull request #33 from r0man/plz
       new  fa6c445271 Fix issue with not handline the plz response correctly 
for streaming
       new  1fb8511867 Use media type only in the llm-request-plz-async call
       new  5623482527 Fix error handling in Gemini and Vertex provider
       new  f42e00265f Add more documentation
       new  2426f827d7 Add process filter error handling
       new  ffbaa5e683 Restore error callbacks
       new  7b2c89f087 Remove invalid media type usage and on-success-raw 
callback
       new  f91d965490 Add a clause to handle process filter errors
       new  669456cba5 Fix Ollama error handling and appending to prompt
       new  c26bf648a8 Increase llm-request-plz-timeout to 2 minutes
       new  ccfe066654 Widen before parsing the response
       new  b438ae7a72 Merge pull request #35 from r0man/plz
       new  7cb3dd4891 Merge branch 'main' into plz
       new  609c8a6834 Fix error message handling
       new  f409841da3 Add ability to cancel processes, which is what plz 
returns
       new  57cf82e4af Update plz-media-type.el
       new  1a40b6d656 Really set reponse body to nil
       new  9a9b0de0df Add llm-request-plz-connect-timeout
       new  c10187a07a Set llm-request-plz-timeout to nil
       new  0cb682aff4 Merge pull request #39 from r0man/plz
       new  29d7c3f6b8 Update media type and event source packages
       new  759f689fae Merge pull request #40 from r0man/plz
       new  5145074ba0 Merge branch 'main' into plz
       new  9d215d7ba2 Enable handling errors on streaming, and add requires
       new  006cd1d6b2 Handle safety related errors in Vertex / Gemini
       new  0fec127d01 Don't always return "Unknown error" in llm-chat-async
       new  8a384aa073 Run handler code via a timer
       new  dda5d1f2d0 Call handler functions only with the event object
       new  54f38fef5b Merge pull request #41 from r0man/plz-timer
       new  2146ba3803 Change some function names and make most of them private 
(#42)
       new  6054da6b58 Plz (#46)
       new  0f25d08218 Merge branch 'plz'
       new  3a2d6962f3 Whitespace fixes
       new  95b9b8b515 Improve error handling in Vertex and Ollama
       new  ece9b1fc82 Remove llm-request and fix requires
       new  6408551c40 Add NEWS entry for the plz migration
       new  478afbcb41 Fix byte compilation errors


Summary of changes:
 NEWS.org              |   5 +
 llm-claude.el         |  43 ++-
 llm-gemini.el         |   1 -
 llm-gpt4all.el        |   1 -
 llm-llamacpp.el       |   1 -
 llm-ollama.el         |  98 ++-----
 llm-openai.el         | 131 ++++------
 llm-provider-utils.el | 112 ++++----
 llm-request-plz.el    | 169 ++++++++++++
 llm-request.el        | 178 -------------
 llm-tester.el         |   2 +-
 llm-vertex.el         |  39 +--
 llm.el                |  11 +-
 plz-event-source.el   | 477 +++++++++++++++++++++++++++++++++
 plz-media-type.el     | 712 ++++++++++++++++++++++++++++++++++++++++++++++++++
 15 files changed, 1548 insertions(+), 432 deletions(-)
 create mode 100644 llm-request-plz.el
 delete mode 100644 llm-request.el
 create mode 100644 plz-event-source.el
 create mode 100644 plz-media-type.el