Rename elmer to ellmer (#237)

To avoid case-insensitive clash with ELMER on bioconductor.

.Rbuildignore

@@ -1,4 +1,4 @@
-^elmer\.Rproj$
+^ellmer\.Rproj$
 ^\.Rproj\.user$
 ^LICENSE\.md$
 ^\.github$

DESCRIPTION

@@ -1,4 +1,4 @@
-Package: elmer
+Package: ellmer
 Title: Call LLM APIs from R
 Version: 0.0.0.9000
 Authors@R: c(
@@ -9,8 +9,8 @@ Authors@R: c(
 Description: A consistent interface for calling LLM APIs. Includes support
     for streaming.
 License: MIT + file LICENSE
-URL: https://elmer.tidyverse.org, https://github.com/tidyverse/elmer
-BugReports: https://github.com/tidyverse/elmer/issues
+URL: https://ellmer.tidyverse.org, https://github.com/tidyverse/ellmer
+BugReports: https://github.com/tidyverse/ellmer/issues
 Imports:
     cli,
     coro (>= 1.1.0),
@@ -55,7 +55,7 @@ Collate:
     'chat.R'
     'content-image.R'
     'content-tools.R'
-    'elmer-package.R'
+    'ellmer-package.R'
     'httr2.R'
     'import-standalone-obj-type.R'
     'import-standalone-purrr.R'

LICENSE

@@ -1,2 +1,2 @@
 YEAR: 2024
-COPYRIGHT HOLDER: elmer authors
+COPYRIGHT HOLDER: ellmer authors

LICENSE.md

@@ -1,6 +1,6 @@
 # MIT License
 
-Copyright (c) 2024 elmer authors
+Copyright (c) 2024 ellmer authors
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

NEWS.md

@@ -1,4 +1,4 @@
-# elmer (development version)
+# ellmer (development version)
 
 * New `chat_vllm()` to chat with models served by vLLM (#140).
 
@@ -28,7 +28,7 @@
 
 * The `echo` can now be one of three values: "none", "text", or "all". If "all", you'll now see both user and assistant turns, and all content types will be printed, not just text. When running in the global environment, `echo` defaults to "text", and when running inside a function it defaults to "none".
 
-* You can now log low-level JSON request/response info by setting `options(elmer_verbosity = 2)`.
+* You can now log low-level JSON request/response info by setting `options(ellmer_verbosity = 2)`.
 
 * `chat$register_tool()` now takes an object created by `Tool()`. This makes it a little easier to reuse tool definitions (#32).
 

R/chat.R

@@ -249,13 +249,13 @@ Chat <- R6::R6Class("Chat",
     },
 
     #' @description Register a tool (an R function) that the chatbot can use.
-    #'   If the chatbot decides to use the function,  elmer will automatically
+    #'   If the chatbot decides to use the function,  ellmer will automatically
     #'   call it and submit the results back.
-    #' 
+    #'
     #'   The return value of the function. Generally, this should either be a
     #'   string, or a JSON-serializable value. If you must have more direct
     #'   control of the structure of the JSON that's returned, you can return a
-    #'   JSON-serializable value wrapped in [base::I()], which elmer will leave
+    #'   JSON-serializable value wrapped in [base::I()], which ellmer will leave
     #'   alone until the entire request is JSON-serialized.
     #' @param tool_def Tool definition created by [tool()].
     register_tool = function(tool_def) {

R/content-image.R

@@ -15,12 +15,12 @@
 #'   the `chat()`, `stream()`, `chat_async()`, or `stream_async()` methods.
 #'
 #' @export
-#' @examplesIf elmer:::openai_key_exists()
+#' @examplesIf ellmer:::openai_key_exists()
 #' chat <- chat_openai(echo = TRUE)
 #' chat$chat(
 #'   "What do you see in these images?",
 #'   content_image_url("https://www.r-project.org/Rlogo.png"),
-#'   content_image_file(system.file("httr2.png", package = "elmer"))
+#'   content_image_file(system.file("httr2.png", package = "ellmer"))
 #' )
 #'
 #' \dontshow{dev.control('enable')}
@@ -144,7 +144,7 @@ content_image_plot <- function(width = 768, height = 768) {
 
   old <- grDevices::dev.cur()
 
-  path <- tempfile("elmer-plot-", fileext = ".png")
+  path <- tempfile("ellmer-plot-", fileext = ".png")
   defer(unlink(path))
 
   grDevices::png(path, width = width, height = height)

R/content.R

@@ -2,11 +2,11 @@
 NULL
 
 #' Format contents into a textual representation
-#' 
+#'
 #' @description
 #' These generic functions can be use to convert [Turn] contents or [Content]
-#' objects into textual representations. 
-#' 
+#' objects into textual representations.
+#'
 #' * `contents_text()` is the most minimal and only includes [ContentText]
 #'   objects in the output.
 #' * `contents_markdown()` returns the text content (which it assumes to be
@@ -15,7 +15,7 @@ NULL
 #' * `contents_html()` returns the text content, converted from markdown to
 #'   HTML with [commonmark::markdown_html()], plus HTML representations of
 #'   images and other content types.
-#' 
+#'
 #' @examples
 #' turns <- list(
 #'   Turn("user", contents = list(
@@ -30,12 +30,12 @@ NULL
 #' if (rlang::is_installed("commonmark")) {
 #'   contents_html(turns[[1]])
 #' }
-#' 
+#'
 #' @param content The [Turn] or [Content] object to be converted into text.
 #'   `contents_markdown()` also accepts [Chat] instances to turn the entire
 #'   conversation history into markdown text.
 #' @param ... Additional arguments passed to methods.
-#' 
+#'
 #' @return A string of text, markdown or HTML.
 #' @export
 contents_text <- new_generic("contents_text", "content")
@@ -52,7 +52,7 @@ contents_markdown <- new_generic("contents_markdown", "content")
 #' Content types received from and sent to a chatbot
 #'
 #' @description
-#' elmer abstracts away differences in the way that different [Provider]s
+#' ellmer abstracts away differences in the way that different [Provider]s
 #' represent various types of content, allowing you to more easily write
 #' code that works with any chatbot.
 #'

R/httr2.R

@@ -153,5 +153,5 @@ cat_with_prefix <- function(prefix, lines) {
 }
 
 log_http_traffic <- function() {
-  getOption("elmer_verbosity", 0L) >= 2L
+  getOption("ellmer_verbosity", 0L) >= 2L
 }

R/provider-bedrock.R

@@ -125,7 +125,7 @@ method(chat_resp_stream, ProviderBedrock) <- function(provider, resp) {
   resp_stream_aws(resp)
 }
 
-# Bedrock -> elmer -------------------------------------------------------------
+# Bedrock -> ellmer -------------------------------------------------------------
 
 method(stream_parse, ProviderBedrock) <- function(provider, event) {
   if (is.null(event)) {
@@ -216,7 +216,7 @@ method(value_turn, ProviderBedrock) <- function(provider, result, has_type = FAL
   Turn(result$output$message$role, contents, json = result, tokens = tokens)
 }
 
-# elmer -> Bedrock -------------------------------------------------------------
+# ellmer -> Bedrock -------------------------------------------------------------
 
 # https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_ContentBlock.html
 method(as_json, list(ProviderBedrock, Turn)) <- function(provider, x) {

R/provider-claude.R

@@ -136,7 +136,7 @@ method(chat_request, ProviderClaude) <- function(provider,
   req
 }
 
-# Claude -> elmer --------------------------------------------------------------
+# Claude -> ellmer --------------------------------------------------------------
 
 method(stream_parse, ProviderClaude) <- function(provider, event) {
   if (is.null(event)) {
@@ -221,7 +221,7 @@ method(value_turn, ProviderClaude) <- function(provider, result, has_type = FALS
   Turn(result$role, contents, json = result, tokens = tokens)
 }
 
-# elmer -> Claude --------------------------------------------------------------
+# ellmer -> Claude --------------------------------------------------------------
 
 method(as_json, list(ProviderClaude, Turn)) <- function(provider, x) {
   if (x@role == "system") {

R/provider-cortex.R

@@ -44,7 +44,7 @@ NULL
 #' @inheritParams chat_openai
 #' @inherit chat_openai return
 #' @family chatbots
-#' @examplesIf elmer:::cortex_credentials_exist()
+#' @examplesIf ellmer:::cortex_credentials_exist()
 #' chat <- chat_cortex(
 #'   model_file = "@my_db.my_schema.my_stage/model.yaml"
 #' )
@@ -137,11 +137,11 @@ method(chat_request, ProviderCortex) <- function(provider,
   req <- req_error(req, body = function(resp) resp_body_json(resp)$message)
 
   # Snowflake uses the User Agent header to identify "parter applications",
-  # so identify requests as coming from "r_elmer" (unless an explicit
+  # so identify requests as coming from "r_ellmer" (unless an explicit
   # partner application is set via the ambient SF_PARTNER environment
   # variable).
   req <- req_user_agent(
-    req, paste0("r_elmer/", utils::packageVersion("elmer"))
+    req, paste0("r_ellmer/", utils::packageVersion("ellmer"))
   )
   if (nchar(Sys.getenv("SF_PARTNER")) != 0) {
     req <- req_user_agent(req, Sys.getenv("SF_PARTNER"))
@@ -158,7 +158,7 @@ method(chat_request, ProviderCortex) <- function(provider,
   req
 }
 
-# Cortex -> elmer --------------------------------------------------------------
+# Cortex -> ellmer --------------------------------------------------------------
 
 method(stream_parse, ProviderCortex) <- function(provider, event) {
   # While undocumented, Cortex seems to mostly follow OpenAI API conventions for
@@ -281,7 +281,7 @@ method(value_turn, ProviderCortex) <- function(provider, result, has_type = FALS
 }
 
 
-# elmer -> Cortex --------------------------------------------------------------
+# ellmer -> Cortex --------------------------------------------------------------
 
 # Cortex supports not only "text" content, but also bespoke "suggestions" and
 # "sql" types.

R/provider-databricks.R

@@ -7,7 +7,7 @@
 #'
 #' Databricks models do not support images, but they do support structured
 #' outputs. Tool calling support is also very limited at present; too limited
-#' for `elmer`'s tool calling features to work properly at all.
+#' for `ellmer`'s tool calling features to work properly at all.
 #'
 #' ## Authentication
 #'
@@ -188,7 +188,7 @@ databricks_token <- function(workspace = databricks_workspace(), token = NULL) {
         paste0("https://", host, "/oidc/v1/token"),
         secret = client_secret,
         auth = "header",
-        name = "elmer-databricks-m2m"
+        name = "ellmer-databricks-m2m"
       ),
       oauth_flow_client_credentials,
       # The "all-apis" scope translates to "everything this service principal
@@ -239,7 +239,7 @@ databricks_token <- function(workspace = databricks_workspace(), token = NULL) {
         "databricks-cli",
         paste0("https://", host, "/oidc/v1/token"),
         auth = "body",
-        name = "elmer-databricks-u2m"
+        name = "ellmer-databricks-u2m"
       ),
       oauth_flow_auth_code,
       flow_params = list(

R/provider-gemini.R

@@ -115,7 +115,7 @@ method(chat_request, ProviderGemini) <- function(provider,
   req
 }
 
-# Gemini -> elmer --------------------------------------------------------------
+# Gemini -> ellmer --------------------------------------------------------------
 
 method(stream_parse, ProviderGemini) <- function(provider, event) {
   if (is.null(event)) {
@@ -169,7 +169,7 @@ method(value_turn, ProviderGemini) <- function(provider, result, has_type = FALS
   Turn("assistant", contents, json = result, tokens = tokens)
 }
 
-# elmer -> Gemini --------------------------------------------------------------
+# ellmer -> Gemini --------------------------------------------------------------
 
 # https://ai.google.dev/api/caching#Content
 method(as_json, list(ProviderGemini, Turn)) <- function(provider, x) {

R/provider-openai.R

@@ -43,7 +43,7 @@ NULL
 #' @family chatbots
 #' @export
 #' @returns A [Chat] object.
-#' @examplesIf elmer:::openai_key_exists()
+#' @examplesIf ellmer:::openai_key_exists()
 #' chat <- chat_openai()
 #' chat$chat("
 #'   What is the difference between a tibble and a data frame?
@@ -143,7 +143,7 @@ method(chat_request, ProviderOpenAI) <- function(provider,
   req
 }
 
-# OpenAI -> elmer --------------------------------------------------------------
+# OpenAI -> ellmer --------------------------------------------------------------
 
 method(stream_parse, ProviderOpenAI) <- function(provider, event) {
   if (is.null(event)) {
@@ -202,7 +202,7 @@ method(value_turn, ProviderOpenAI) <- function(provider, result, has_type = FALS
   Turn(message$role, content, json = result, tokens = tokens)
 }
 
-# elmer -> OpenAI --------------------------------------------------------------
+# ellmer -> OpenAI --------------------------------------------------------------
 
 method(as_json, list(ProviderOpenAI, Turn)) <- function(provider, x) {
   if (x@role == "system") {

R/provider-vllm.R

@@ -57,7 +57,7 @@ chat_vllm_test <- function(...) {
 ProviderVllm <- new_class(
   "ProviderVllm",
   parent = ProviderOpenAI,
-  package = "elmer",
+  package = "ellmer",
 )
 
 # Just like OpenAI but no strict

R/tools-def-auto.R

@@ -74,7 +74,7 @@ create_tool_def <- function(topic,
 
   topic_str <- format(expr)
 
-  tool_prompt <- readLines(system.file("tool_prompt.md", package = "elmer"), warn = FALSE)
+  tool_prompt <- readLines(system.file("tool_prompt.md", package = "ellmer"), warn = FALSE)
   tool_prompt <- paste(tool_prompt, collapse = "\n")
 
   payload <- paste0(

R/tools-def.R

@@ -19,7 +19,7 @@ NULL
 #'   function. Each element should be created by a [`type_*()`][type_boolean]
 #'   function.
 #' @export
-#' @examplesIf elmer:::openai_key_exists()
+#' @examplesIf ellmer:::openai_key_exists()
 #'
 #' # First define the metadata that the model uses to figure out when to
 #' # call the tool

R/turns.R

@@ -12,7 +12,7 @@ NULL
 #'
 #' Note that a call to `$chat()` and related functions may result in multiple
 #' user-assistant turn cycles. For example, if you have registered tools,
-#' elmer will automatically handle the tool calling loop, which may result in
+#' ellmer will automatically handle the tool calling loop, which may result in
 #' any number of additional cycles. Learn more about tool calling in
 #' `vignette("tool-calling")`.
 #'
@@ -21,7 +21,7 @@ NULL
 #' @param json The serialized JSON corresponding to the underlying data of
 #'   the turns. Currently only provided for assistant.
 #'
-#'   This is useful if there's information returned by the provider that elmer
+#'   This is useful if there's information returned by the provider that ellmer
 #'   doesn't otherwise expose.
 #' @param tokens A numeric vector of length 2 representing the number of
 #'   input and output tokens (respectively) used in this turn. Currently

R/types.R

@@ -4,7 +4,7 @@ NULL
 #' Type definitions for function calling and structured data extraction.
 #'
 #' These S7 classes are provided for use by package devlopers who are
-#' extending elmer. In every day use, use [type_boolean()] and friends.
+#' extending ellmer. In every day use, use [type_boolean()] and friends.
 #'
 #' @name Type
 #' @inheritParams type_boolean

R/utils-coro.R

@@ -49,7 +49,7 @@ async_method <- function(fn) {
 # upon first call. This is necessary because the transformation needs to be done
 # not at package build time, but after package load time.
 #
-# Elsewhere in elmer, we use rlang::on_load to defer the transformation of
+# Elsewhere in ellmer, we use rlang::on_load to defer the transformation of
 # generators until after package load time. We can't do that for R6 methods
 # because nesting R6 class definitions inside of rlang::on_load causes roxygen2
 # to get confused.