-
Notifications
You must be signed in to change notification settings - Fork 6.5k
fix: Table Optimization #5319
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fix: Table Optimization #5319
Conversation
TestGru AssignmentSummary
Tip You can |
Preview mcp_server Image:
|
Preview sandbox Image:
|
Preview fastgpt Image:
|
row.map((cell) => (typeof cell === 'string' ? cell.replace(/[\n\r]/g, ' ') : cell)) | ||
) | ||
})); | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
数据类型判断重复,肯定是 string。并且不能把换行换成空格,语义不对。
而且应该在原 replace 替换即可,不应该增加计算复杂度
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
不换成空格的话,可以换成“<br”>和“\\n”吗,还是说有别的方法
@@ -468,10 +479,10 @@ export const splitText2Chunks = (props: SplitProps): SplitResponse => { | |||
|
|||
const splitResult = splitWithCustomSign.map((item) => { | |||
if (strIsMdTable(item)) { | |||
return markdownTableSplit(props); | |||
return markdownTableSplit({ ...props, text: item, tableChunkOverflowRatio: 1.2 }); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
这里为啥×1.2?
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
const { chunks: tableChunks } = markdownTableSplit({ | ||
text: currentText, | ||
chunkSize, | ||
tableChunkOverflowRatio: 1.2 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
不用加参数。直接把 props 里的 chunksize×1.2 就好了
@@ -31,7 +31,7 @@ export const readXlsxRawText = async ({ | |||
| ${header.map(() => '---').join(' | ')} | | |||
${csvArr | |||
.slice(1) | |||
.map((row) => `| ${row.map((item) => item.replace(/\n/g, '\\n')).join(' | ')} |`) | |||
.map((row) => `| ${row.map((item) => item.replace(/[\n\r]/g, ' ')).join(' | ')} |`) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
这里不能都改成空格,语义有问题,都要改成转义符
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
改成转义字符的话'\n'和'<br>'都会错位,之前不是聊了说改空格不影响ai嘛,改成转义符的话怎么改呢
* feat: table test: 1 * feat: table test: 2 * feat: table test: 3 * feat: table test: 4 * feat: table test : 5 把maxSize改回chunkSize * feat: table test : 6 都删了,只看maxSize * feat: table test : 7 恢复初始,接下来删除标签功能 * feat: table test : 8 删除标签功能 * feat: table test : 9 删除标签功能成功 * feat: table test : 10 继续调试,修改trainingStates * feat: table test : 11 修改第一步 * feat: table test : 12 修改第二步 * feat: table test : 13 修改了HtmlTable2Md * feat: table test : 14 修改表头分块规则 * feat: table test : 15 前面表格分的太细了 * feat: table test : 16 改着改着表头又不加了 * feat: table test : 17 用CUSTOM_SPLIT_SIGN不行,重新改 * feat: table test : 18 表头仍然还会多加,但现在分块搞的合理了终于 * feat: table test : 19 还是需要搞好表头问题,先保存一下调试情况 * feat: table test : 20 调试结束,看一下replace有没有问题,没问题就pr * feat: table test : 21 先把注释删了 * feat: table test : 21 注释replace都改了,下面切main分支看看情况 * feat: table test : 22 修改旧文件 * feat: table test : 23 修改测试文件 * feat: table test : 24 xlsx表格处理 * feat: table test : 25 刚才没保存先com了 * feat: table test : 26 fix * feat: table test : 27 先com一版调试 * feat: table test : 28 试试放format2csv里 * feat: table test : 29 xlsx解决 * feat: table test : 30 tablesplit解决 * feat: table test : 31 * feat: table test : 32
* perf: system toolset & mcp (#5200) * feat: support system toolset * fix: type * fix: system tool config * chore: mcptool config migrate * refactor: mcp toolset * fix: fe type error * fix: type error * fix: show version * chore: support extract tool's secretInputConfig out of inputs * chore: compatible with old version mcp * chore: adjust * deps: update dependency @fastgpt-skd/plugin * fix: version * fix: some bug (#5316) * chore: compatible with old version mcp * fix: version * fix: compatible bug * fix: mcp object params * fix: type error * chore: update test cases * chore: remove log * fix: toolset node name * optimize app logs sort (#5310) * log keys config modal * multiple select * api * fontsize * code * chatid * fix build * fix * fix component * change name * log keys config * fix * delete unused * fix * perf: log code * perf: send auth code modal enter press * fix log (#5328) * perf: mcp toolset comment * perf: log ui * remove log (#5347) * doc * fix: action * remove log * fix: Table Optimization (#5319) * feat: table test: 1 * feat: table test: 2 * feat: table test: 3 * feat: table test: 4 * feat: table test : 5 把maxSize改回chunkSize * feat: table test : 6 都删了,只看maxSize * feat: table test : 7 恢复初始,接下来删除标签功能 * feat: table test : 8 删除标签功能 * feat: table test : 9 删除标签功能成功 * feat: table test : 10 继续调试,修改trainingStates * feat: table test : 11 修改第一步 * feat: table test : 12 修改第二步 * feat: table test : 13 修改了HtmlTable2Md * feat: table test : 14 修改表头分块规则 * feat: table test : 15 前面表格分的太细了 * feat: table test : 16 改着改着表头又不加了 * feat: table test : 17 用CUSTOM_SPLIT_SIGN不行,重新改 * feat: table test : 18 表头仍然还会多加,但现在分块搞的合理了终于 * feat: table test : 19 还是需要搞好表头问题,先保存一下调试情况 * feat: table test : 20 调试结束,看一下replace有没有问题,没问题就pr * feat: table test : 21 先把注释删了 * feat: table test : 21 注释replace都改了,下面切main分支看看情况 * feat: table test : 22 修改旧文件 * feat: table test : 23 修改测试文件 * feat: table test : 24 xlsx表格处理 * feat: table test : 25 刚才没保存先com了 * feat: table test : 26 fix * feat: table test : 27 先com一版调试 * feat: table test : 28 试试放format2csv里 * feat: table test : 29 xlsx解决 * feat: table test : 30 tablesplit解决 * feat: table test : 31 * feat: table test : 32 * perf: table split * perf: mcp old version compatibility (#5342) * fix: system-tool secret inputs * fix: rewrite runtime node i18n for system tool * perf: mcp old version compatibility * fix: splitPluginId * fix: old mcp toolId * fix: filter secret key * feat: support system toolset activation * chore: remove log * perf: mcp update * perf: rewrite toolset * fix:delete variable id (#5335) * perf: variable update * fix: multiple select ui * perf: model config move to plugin * fix: var conflit * perf: variable checker * Avoid empty number * update doc time * fix: test * fix: mcp object * update count app * update count app --------- Co-authored-by: Finley Ge <[email protected]> Co-authored-by: heheer <[email protected]> Co-authored-by: heheer <[email protected]> Co-authored-by: colnii <[email protected]> Co-authored-by: dreamer6680 <[email protected]>
No description provided.