Select latest envname if match not available, with a bunch of messages

R/import-check.R

@@ -29,12 +29,31 @@ import_check <- function(x, envname) {
     env_loaded <- env_python(envname) == py_exe()
   }
 
-  inst <- paste0(
-    " {.run pysparklyr::install_pyspark(",
-    "envname = \"{envname}\")}"
-  )
+  inst <- NULL
+
+  if(substr(envname, 1, 22) == "r-sparklyr-databricks-") {
+    inst <- paste0(
+      " {.run pysparklyr::install_databricks(",
+      "envname = \"{envname}\")}"
+    )
+  }
+
+  if(substr(envname, 1, 19) == "r-sparklyr-pyspark-") {
+    inst <- paste0(
+      " {.run pysparklyr::install_pyspark(",
+      "envname = \"{envname}\")}"
+    )
+  }
+
+  msg_install <- NULL
+  msg_restart <- NULL
+  if(!is.null(inst)) {
+    # msg_install <- paste("{.header - Use} ", inst, "{.header to install.}")
+    # msg_restart <- paste("- Restart your R session, and run:", inst)
+  }
 
   if (inherits(out, "try-error")) {
+    cli_alert_danger(glue("`reticulate` error:\n {out[[1]]}"))
     if (env_found) {
       if (env_loaded) {
         # found & loaded
@@ -43,8 +62,8 @@ import_check <- function(x, envname) {
             "{.emph '{x}' }{.header is not available in the }",
             "{.emph '{envname}' }{.header Python environment.}"
           ),
-          paste("{.header - Use}", inst, "{.header to install.}")
-        ))
+          msg_install
+        ), call = NULL)
       } else {
         cli_abort(c(
           "{.emph '{x}' }{.header is not available current Python environment.}",
@@ -56,14 +75,14 @@ import_check <- function(x, envname) {
             "{.header - Restart your R session, and avoid",
             " initializing Python before using} {.emph '{x}'}"
           )
-        ))
+        ), call = NULL)
       }
     } else {
       cli_abort(c(
         "{.emph '{x}' }{.header is not available current Python environment.}",
         paste("- The {.emph '{envname}'} Python environment is not installed."),
-        paste("- Restart your R session, and run:", inst)
-      ))
+        msg_restart
+      ), call = NULL)
     }
   } else {
     if (env_loaded) {

R/install.R

@@ -107,7 +107,7 @@ install_environment <- function(
   ver_name <- version_prep(version)
 
   if(version == ver_name) {
-    version <- paste(version, ".*")
+    version <- paste0(version, ".*")
   }
 
   if (is.null(envname)) {

R/spark-connect.R

@@ -81,7 +81,23 @@ py_spark_connect <- function(master,
       dbr <- version_prep(dbr_version)
     }
 
-    envname <- glue("r-sparklyr-databricks-{dbr}")
+    env_base <- "r-sparklyr-databricks-"
+    envname <- glue("{env_base}{dbr}")
+    envs <- find_environments(env_base)
+    matched <- envs[envs == envname]
+    if(length(matched) == 0) {
+      envname <- envs[[1]]
+      cli_div(theme = cli_colors())
+      cli_alert_warning(paste(
+        "{.header A Python environment with a matching version was not found}",
+        "* {.header Will attempt connecting using }{.emph '{envname}'}",
+        paste0("* {.header To install the proper Python environment use:}",
+               " {.run pysparklyr::install_databricks(version = \"{dbr}\")}"
+               ),
+        sep = "\n"
+        ))
+      cli_end()
+    }
 
     db <- import_check("databricks.connect", envname)
     remote <- db$DatabricksSession$builder$remote(
@@ -285,3 +301,13 @@ cluster_dbr_info <- function(cluster_id,
     req_perform() %>%
     resp_body_json()
 }
+
+
+find_environments <- function(x) {
+  conda_names <- conda_list()$name
+  ve_names <- virtualenv_list()
+  all_names <- c(ve_names, conda_names)
+  sub_names <- substr(all_names, 1, nchar(x))
+  matched <- all_names[sub_names == x]
+  sort(matched, decreasing = FALSE)
+}