xbhh XLSX查看器

Somewhere between "why would you ever need this" and "I now use it every week" sits the XBHH XLSX viewer. It reads an .xlsx file, renders the sheet as a formatted text table right in the node, and hands you the content plus row and column counts as outputs. If your prompts, LoRA lists, or batch parameters live in a spreadsheet - which a surprising number of people's actually do - this lets the graph itself read them.

The classic setup: a spreadsheet where each row is a prompt variation or a character spec, you pick one (or just eyeball the sheet), and the node turns spreadsheet rows into workflow inputs. It also earns its keep as a lightweight inspection tool when you're handed a config workbook and don't want to alt-tab out of ComfyUI to read it.

How it works

It's a thin wrapper around openpyxl - the only real Python dependency in the entire XBHH kit, and this node is why it exists. On execution it loads the workbook in read-only mode, picks the requested sheet (or the first one if sheet_name is blank), reads up to max_rows rows, and formats them into an aligned pipe-separated table. It caps column width at 20 characters with a .. truncation marker so huge cells don't turn the output into a wall. Values are read with data_only=True, which means you get computed results rather than formulas.

The inputs and outputs that matter

• file_path - the full path to the .xlsx file. Required, and it's a path, not a dropdown - there's no file picker.
• sheet_name (optional) - pick a specific sheet by name; blank means the first/active sheet.
• max_rows (optional) - cap on how many rows get read and rendered, default 100, max 10000. Raise it for big sheets, lower it for speed on enormous ones.

Three outputs: content (STRING - the whole rendered table), row_count and col_count (INTs - the true totals from the sheet, not just what you capped). Those counts are the sleeper feature: wire row_count into iteration logic or a batch-size calculator and the spreadsheet drives the workflow.

Installing it

This is the node that makes the kit's install step non-optional:

cd ComfyUI/custom_nodes
git clone https://github.com/xbhh123456/Comfyui-xbhh-main.git
cd Comfyui-xbhh-main
pip install -r requirements.txt

That requirements.txt is essentially just openpyxl>=3.1.0. Skip the pip step and this node will happily tell you exactly what's missing. ComfyUI Manager users can search xbhh-lora and Manager handles the dependency.

Common issues

• "❌ 需要安装 openpyxl 库" - you skipped the pip install, or you installed the pack into an environment where requirements.txt never ran. pip install openpyxl in the same environment as ComfyUI, restart, done.
• "❌ 仅支持 .xlsx 格式文件" - it's .xlsx only. Old .xls files from Excel 97–2003 are rejected; re-save them as .xlsx first.
• "❌ 文件不存在" - the node takes a real filesystem path, not a ComfyUI-relative path. On Windows, mind that your path string might need C:/Users/... forward slashes depending on how you pasted it.
• The table view is read-only by design - this is a viewer, not an editor. If you want to modify sheets from a workflow, this isn't that node.

One honest note: this node is a specialty tool. You'll either have a spreadsheet-driven workflow and consider it essential, or never touch it. Both are fine - but if you ever find yourself hand-copying rows from Excel into a text widget, you now know the better way.