Feature/subfolders in outputs

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: pm
 Title: Project and Analysis Manager
-Version: 0.1.10
+Version: 0.1.11
 Authors@R: 
     person("Alon", "Alexander", , "alon008@gmail.com", role = c("aut", "cre"))
 Description: Enforces and supports a standardized folder structure for research

NEWS.md

@@ -1,3 +1,11 @@
+# pm 0.1.11
+
+## New Features
+
+- Support subfolders in `PMAnalysis$get_output_path`
+- Support subfolders in `PMAnalysis$list_outputs`
+- Support subfolders in `get_artifact` (both `PMAnalysis` and `PmProject`)
+
 # pm 0.1.10
 
 ## New Features

R/analysis.R

@@ -241,31 +241,23 @@ PMAnalysis <- R6Class("PMAnalysis",
       }
 
       # Get all files in the directory
-      files <- list.files(outputs_dir, full.names = TRUE)
+      files <- list.files(outputs_dir, full.names = FALSE, recursive = TRUE)
       if (length(files) == 0) {
         return(list())
       }
 
-      # Filter to only files (not directories) using vectorized approach
-      file_info <- file.info(files)
-      is_file <- !is.na(file_info$isdir) & !file_info$isdir
-      files_only <- files[is_file]
-
-      if (length(files_only) == 0) {
-        return(list())
-      }
-
       # Create PMData objects using lapply
-      file_ids <- tools::file_path_sans_ext(basename(files_only))
-      file_paths <- normalizePath(files_only, mustWork = FALSE)
+      file_ids <- gsub("\\", "/", tools::file_path_sans_ext(files), fixed = TRUE)
+      file_paths <- normalizePath(file.path(outputs_dir, files), mustWork = FALSE)
 
-      lapply(seq_along(files_only), function(i) {
+      lapply(seq_along(files), function(i) {
         PMData$new(id = file_ids[i], path = file_paths[i])
       })
     },
 
     #' @description
     #' Get output path for a file, returning a PMData object.
+    #' Supports also subfolders using both unix-style and windows-style delimeteres ("/" and "\\").
     #'
     #' @param name Character. Name of the output file (with or without extension).
     #' @param type Character. Optional type of output (table, object, image, figure, parquet, csv).
@@ -291,6 +283,11 @@ PMAnalysis <- R6Class("PMAnalysis",
     #' intermediate <- analysis$get_output_path("temp_data", type = "table", intermediate = TRUE)
     #' intermediate$id    # "temp_data"
     #' intermediate$path  # full path to temp_data.parquet in intermediate/
+    #'
+    #' # Get output path with nested folders
+    #' output2 <- analysis$get_output_path("unique\\complex\\structure.rds")
+    #' output2$id   # "unique/complex/structure"
+    #' output2$path
     get_output_path = function(name, type = NULL, intermediate = FALSE) {
       # Store original name for ID (without extension)
       original_name <- name
@@ -332,7 +329,9 @@ PMAnalysis <- R6Class("PMAnalysis",
 
       folder <- if (intermediate) constants$ANALYSIS_INTERMEDIATE_DIR else constants$ANALYSIS_OUTPUT_DIR
 
-      full_path <- normalizePath(file.path(self$path, folder, name), mustWork = FALSE)
+      name <- strsplit(name, "\\\\|/")
+      full_path_raw <- do.call(file.path, as.list(c(self$path, folder, unlist(name))))
+      full_path <- normalizePath(full_path_raw, mustWork = FALSE)
 
       PMData$new(id = id, path = full_path)
     },

R/data.R

@@ -31,7 +31,7 @@ PMData <- R6Class("PMData",
       chk::chk_scalar(path)
       chk::chk_character(path)
       self$id <- id
-      self$path <- path
+      self$path <- normalizePath(path, mustWork = FALSE)
     },
 
     #' @description
@@ -101,6 +101,9 @@ PMData <- R6Class("PMData",
     #'   data_rdata$write(obj1, obj2, obj3 = 42)
     #' })
     write = function(x, ...) {
+      # Create the folder in case it doesn't exist
+      dir.create(dirname(self$path), showWarnings = FALSE, recursive = TRUE)
+
       # For RData files, we need   to preserve object names from the original call
       ext <- tolower(tools::file_ext(self$path))
       if (ext %in% c("rdata", "rda")) {
@@ -129,7 +132,7 @@ PMData <- R6Class("PMData",
             obj_names <- c(obj_names, obj_name)
           }
         }
-        
+
         # Call pm_write_file with explicit object names
         pm_write_file(self$path, x, ..., object_names = obj_names)
       } else {

R/data_utils.R

@@ -744,4 +744,3 @@ pm_write_file <- function(file, x, ..., object_names = NULL) {
 
   invisible(file)
 }
-

R/project.R

@@ -275,6 +275,10 @@ PMProject <- R6Class("PMProject",
       chk::chk_scalar(id)
       chk::chk_character(id)
 
+      # Normalize (subfolder) id
+      prev_id = id
+      id <- gsub("\\", "/", id, fixed = TRUE)
+
       # Determine which analyses to search
       if (!is.null(analysis_name)) {
         chk::chk_scalar(analysis_name)

docs/articles/file-formats.md

@@ -145,7 +145,7 @@ head(tsv_data)
 # Read the RDS object (automatically detects RDS format)
 model <- analysis$get_artifact("model")$read()
 str(model, max.level = 2)
-#> <environment: 0x1228fe2e8>
+#> <environment: 0x10d6cec18>
 
 # Read RData (returns an environment)
 rdata_env <- analysis$get_artifact("model_rdata")$read()