From 21174836dec9754fcd66034142b723de350aaa6d Mon Sep 17 00:00:00 2001
From: gaozhaochen <158975971@qq.com>
Date: Wed, 26 Nov 2025 16:54:55 +0800
Subject: [PATCH] =?UTF-8?q?update:=20=E8=87=AA=E5=8A=A8=E5=8A=A0=E8=BD=BDx?=
 =?UTF-8?q?lsx=E7=9A=84=E5=85=A8=E9=83=A8sheet=E6=95=B0=E6=8D=AE?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .../inst/app/tools/data/init_load_data.R      |  67 +++++++-----
 radiant.data/inst/app/tools/data/manage.R     | 101 +++++++++++++-----
 2 files changed, 113 insertions(+), 55 deletions(-)

diff --git a/radiant.data/inst/app/tools/data/init_load_data.R b/radiant.data/inst/app/tools/data/init_load_data.R
index a7bed85..a31f485 100644
--- a/radiant.data/inst/app/tools/data/init_load_data.R
+++ b/radiant.data/inst/app/tools/data/init_load_data.R
@@ -8,38 +8,36 @@ observe({
   dataset_id <- query[['datasetId']]
   token <- query[['token']]
 
+  # 初始化一个 session 级别的变量来记录已加载的 ID，防止刷新时的重复请求
+  if (is.null(session$userData$loaded_datasets)) {
+    session$userData$loaded_datasets <- c()
+  }
+
   # 2. 仅当 ID 和 Token 均存在时执行
   if (!is.null(dataset_id) && !is.null(token)) {
 
-    # 定义 Radiant 内部使用的数据集名称 (例如: data_10086)
-    # 如果 URL 传了 name 参数就用 name，否则用 id 拼接
-    ds_name <- if (!is.null(query[['name']])) query[['name']] else paste0("data_", dataset_id)
-
-    # 3. 检查数据是否已存在 (防止重复加载)
-    if (is.null(r_data[[ds_name]])) {
+    # 3. [修改点] 检查 ID 是否已在本次会话中加载过
+    # 因为现在数据集名称是动态的 Sheet 名，下载前无法预知，所以改为判断 ID
+    if (!(dataset_id %in% session$userData$loaded_datasets)) {
 
       withProgress(message = '正在从业务系统同步数据...', value = 0.2, {
 
-        # 4. 获取环境变量中的 API 基地址
-        # api_base <- Sys.getenv("HOST_API_BASE","http://127.0.0.1:11999")
-        api_base <- Sys.getenv("HOST_API_BASE","https://ds.cixincloud.com/data-search-api")
+        # 4. 获取环境变量
+        api_base <- Sys.getenv("HOST_API_BASE", "https://ds.cixincloud.com/data-search-api")
 
         # 5. 拼接完整 API 路径
-        # target_url <- paste0(api_base, "/disease-data/data/export/apply/apply/case?applyId=", dataset_id)
         target_url <- paste0(api_base, "/research-project/generate-project-dataset/", dataset_id)
 
-        # 6. 创建临时文件 (明确 .xlsx 后缀)
+        # 6. 创建临时文件
         tmp_file <- tempfile(fileext = ".xlsx")
 
         tryCatch({
           incProgress(0.3, detail = "正在鉴权并下载...")
 
-          # 7. 发起带 Token 的 HTTP 请求
+          # 7. HTTP 请求
           response <- httr::POST(
             url = target_url,
-            # 添加 Bearer Token (或根据你的接口要求修改 Header)
             httr::add_headers(authentication = paste(token)),
-            # 将结果写入磁盘
             httr::write_disk(tmp_file, overwrite = TRUE)
           )
 
@@ -47,34 +45,47 @@ observe({
           print(response$request)
           message("===================")
 
-          # 检查 HTTP 状态码
           if (httr::status_code(response) != 200) {
-            stop(paste("下载失败，请手动导入，HTTP状态码:", httr::status_code(response)))
+            stop(paste("下载失败，HTTP状态码:", httr::status_code(response)))
           }
 
           incProgress(0.7, detail = "解析并导入 Radiant...")
 
-          # 8. 复用 Radiant 核心加载函数 (manage_ui.R 中定义)
-          # 这会自动完成读取、转因子、生成R代码、注册到下拉框等所有动作
+          # 8. 调用核心加载函数
+          # 注意：这里的 fname 仅用于扩展名识别，实际数据集名称由 Sheet 名决定
           load_user_data(
-            fname = paste0(ds_name, ".xlsx"), # 虚拟文件名
-            uFile = tmp_file,                 # 实际文件路径
+            fname = "downloaded_data.xlsx",
+            uFile = tmp_file,
             ext = "xlsx",
-            xlsx_sheet = 1,
             xlsx_header = TRUE,
             man_str_as_factor = TRUE
           )
 
-          # 9. 界面联动：选中数据并跳转到视图
-          updateSelectInput(session, "dataset", selected = ds_name)
-          updateTabsetPanel(session, "nav_radiant", selected = "Data")
-          updateTabsetPanel(session, "tabs_data", selected = "View") # 或者 "Visualize"
+          # 9. 界面联动：获取刚才加载的 Sheet 名称并选中第一个
+          # 必须再次读取 Sheet 列表并做 make.names 处理，才能匹配到 r_data 中的 Key
+          loaded_sheets <- try(readxl::excel_sheets(tmp_file), silent = TRUE)
+
+          if (!inherits(loaded_sheets, "try-error") && length(loaded_sheets) > 0) {
 
-          showNotification(paste("数据集", ds_name, "加载成功！"), type = "message")
+            # 必须和 load_user_data 里的逻辑一致，使用 make.names 清洗
+            valid_names <- make.names(loaded_sheets)
+            first_sheet_name <- valid_names[1] # 默认选第一个
+
+            # 标记为已加载
+            session$userData$loaded_datasets <- c(session$userData$loaded_datasets, dataset_id)
+
+            # 更新下拉框并选中第一个 Sheet
+            updateSelectInput(session, "dataset", selected = first_sheet_name)
+            updateTabsetPanel(session, "nav_radiant", selected = "Data")
+            updateTabsetPanel(session, "tabs_data", selected = "View")
+
+            showNotification(paste0("成功加载 ", length(valid_names), " 个数据集 (ID: ", dataset_id, ")"), type = "message")
+          } else {
+            showNotification("文件下载成功，但未发现有效的 Sheet", type = "warning")
+          }
 
         }, error = function(e) {
-          showNotification(paste("数据同步失败，请手动导入数据:", e$message), type = "error", duration = 10)
-          #调试打印 print(e)
+          showNotification(paste("数据同步失败:", e$message), type = "error", duration = 10)
         })
       })
     }
diff --git a/radiant.data/inst/app/tools/data/manage.R b/radiant.data/inst/app/tools/data/manage.R
index 583b24a..3cfea63 100644
--- a/radiant.data/inst/app/tools/data/manage.R
+++ b/radiant.data/inst/app/tools/data/manage.R
@@ -72,13 +72,13 @@ load_user_data <- function(fname, uFile, ext, header = TRUE,
     fext %in% c("xls", "xlsx") ~ "xlsx",
     TRUE ~ ext
   )
-  
+
 
   ## objname is used as the name of the data.frame, make case insensitive
   objname <- sub(glue("\\.{ext}$"), "", filename, ignore.case = TRUE)
 
   ## if ext isn't in the filename nothing was replaced and so ...
-  if (objname == filename && !fext %in% c("xls", "xlsx")) { 
+  if (objname == filename && !fext %in% c("xls", "xlsx")) {
     ret <- glue(i18n$t('#### The filename extension "{fext}" does not match the specified \\
     file-type "{ext}". Please specify the file type you are trying to upload'))
     upload_error_handler(objname, ret)
@@ -142,35 +142,77 @@ load_user_data <- function(fname, uFile, ext, header = TRUE,
     }
   }else if (ext == "xlsx") {
     if (!requireNamespace("readxl", quietly = TRUE)) {
-      ret <- i18n$t("#### 读取xlsx文件需要readxl包")
-      upload_error_handler(objname, ret)
-    } else {
-      # 用readxl读取xlsx
-      robj <- try(readxl::read_excel(
+      upload_error_handler(objname, i18n$t("#### 需要安装 readxl 包"))
+      return()
+    }
+
+    # 1. 获取所有 Sheet
+    all_sheets <- try(readxl::excel_sheets(path = uFile), silent = TRUE)
+
+    if (inherits(all_sheets, "try-error")) {
+      upload_error_handler(objname, i18n$t("#### 无法解析 Excel 文件结构"))
+      return()
+    }
+
+    loaded_count <- 0
+
+    # 2. 循环处理
+    for (sheet_name in all_sheets) {
+
+      # 2.1 准备合法的变量名
+      clean_name <- make.names(sheet_name)
+      if (clean_name == "") clean_name <- paste0("Sheet_", sample(1000, 1))
+
+      # 2.2 读取数据
+      raw_data <- try(readxl::read_excel(
         path = uFile,
-        sheet = xlsx_sheet,    # 对应UI的“工作表索引”
-        col_names = xlsx_header# 对应UI的“第一行为表头”
+        sheet = sheet_name,
+        col_names = xlsx_header
       ), silent = TRUE)
-      
-      if (inherits(robj, "try-error")) {
-        upload_error_handler(objname, i18n$t("#### 读取xlsx文件失败，请检查文件是否损坏或格式正确"))
-      } else {
-        # 转换为data.frame并处理因子
-        r_data[[objname]] <- as.data.frame(robj, stringsAsFactors = FALSE) %>%
-          {if (man_str_as_factor) radiant.data::to_fct(.) else .}
-        # 生成R代码
-        cmd <- glue('
-        {objname} <- readxl::read_excel(
-          {pp$rpath},
-          sheet = {xlsx_sheet},
-          col_names = {xlsx_header}
-        ) %>%
-        as.data.frame(stringsAsFactors = FALSE)
-        {if (man_str_as_factor) paste0(objname, " <- radiant.data::to_fct(", objname, ")") else ""}
-        register("{objname}")
+
+      # 2.3 卫语句：跳过无效数据（空表或读取错误）
+      if (inherits(raw_data, "try-error") || !is.data.frame(raw_data) || ncol(raw_data) == 0) {
+        next
+      }
+
+      # 2.4 [关键] 立即执行 Radiant 标准数据清洗流程
+      # 必须在循环内做，因为我们不会走到函数底部的通用处理逻辑了
+      try({
+        # 清洗列名
+        colnames(raw_data) <- radiant.data::fix_names(colnames(raw_data))
+        # 转换因子
+        if (man_str_as_factor) raw_data <- radiant.data::to_fct(raw_data)
+      }, silent = TRUE)
+
+      # 2.5 入库
+      r_data[[clean_name]] <- as.data.frame(raw_data, stringsAsFactors = FALSE)
+
+      # 2.6 生成并注册复现代码 (Reproducibility)
+      # 注意：这里直接注册 register，不需要 cmd 变量在外部累加
+      curr_cmd <- glue::glue('
+        {clean_name} <- readxl::read_excel("{uFile}", sheet = "{sheet_name}", col_names = {xlsx_header}) %>%
+          as.data.frame(stringsAsFactors = FALSE) %>%
+          fix_names()
+        {if (man_str_as_factor) paste0(clean_name, " <- radiant.data::to_fct(", clean_name, ")") else ""}
+        register("{clean_name}")
       ')
+
+      # 2.7 更新系统元数据 (Mimic bottom logic)
+      if (exists(clean_name, envir = r_data) && !bindingIsActive(as.symbol(clean_name), env = r_data)) {
+        shiny::makeReactiveBinding(clean_name, env = r_data)
       }
-    } 
+      r_info[[glue("{clean_name}_lcmd")]] <- curr_cmd
+      r_info[["datasetlist"]] <- unique(c(clean_name, r_info[["datasetlist"]]))
+
+      loaded_count <- loaded_count + 1
+    }
+
+    if (loaded_count == 0) {
+      upload_error_handler(objname, i18n$t("#### 文件中未发现有效的表格数据"))
+    }
+
+    # 直接返回
+    return()
   }else if (ext %in% c("tsv", "csv", "txt")) {
     r_data[[objname]] <- load_csv(
       uFile,
@@ -197,6 +239,11 @@ load_user_data <- function(fname, uFile, ext, header = TRUE,
     upload_error_handler(objname, ret)
   }
 
+  # ===========================================================================
+  # 下面的代码只为单数据集流程（CSV, RDS, parquet）服务
+  # XLSX 流程已经在上面 return() 了，不会执行到这里
+  # ===========================================================================
+
   if (exists(objname, envir = r_data) && !bindingIsActive(as.symbol(objname), env = r_data)) {
     shiny::makeReactiveBinding(objname, env = r_data)
   }
-- 
2.22.0