# Sending excel file as response is slow

**URL:** <https://discourse.julialang.org/t/sending-excel-file-as-response-is-slow/116705>\
**Category:** Performance\
**Tags:** dataframes, csv, xlsx\
**Created:** [July 6, 2024, 8:47pm UTC](https://discourse.julialang.org/t/sending-excel-file-as-response-is-slow/116705 "2024-07-06T20:47:14Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![nhz2](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/nhz2/32/44428_2.png) [@nhz2](https://discourse.julialang.org/u/nhz2)\
**Post date:** [July 7, 2024, 2:21pm UTC](https://discourse.julialang.org/t/sending-excel-file-as-response-is-slow/116705/4 "2024-07-07T14:21:56Z")

</div>

Thanks for the example data, I did some profiling with StatProfilerHTML.jl and for me, about 80% of the time is spent calling [ZipFile.jl/src/ZipFile.jl at a599b0aac5d17403fdbbc4ea63612be299cf8417 · fhs/ZipFile.jl · GitHub](https://github.com/fhs/ZipFile.jl/blob/a599b0aac5d17403fdbbc4ea63612be299cf8417/src/ZipFile.jl#L677)  
So this seems to be the same issue you were having in [File zipping taking longer for large files](https://discourse.julialang.org/t/file-zipping-taking-longer-for-large-files/115963/1)

Here is what I profiled:

```julia
using CSV
using DataFrames
using XLSX
using StatProfilerHTML

function gen_fake_csv(path::String, nlines::Int)
    header = "REPORT_DT,COMPANY_ID,NUM,DPRT,DPRT_TML,ARRV,ARRV_TML,MARKET,ENTITY,HUB,SUBFT,FLT,DEPS,BHRS,MILES,CAP,MILLIONS,SOURCE,YYYYMM,REPORT_YYYYMM,categ"
    line = "7/1/2024,XY,1,ABC,1234,ABC,1234,ABCDFE,XYZ,ABC,12A,789,1,16.3333,9876,456,2.16959,7/1/2024,202407,202407,787"
    open(path; write=true) do f
        println(f, header)
        join(f, Iterators.repeated(line, nlines), "\n")
    end
end

function create_excel_file(df)
    # Create an in-memory buffer for the Excel file
    io = IOBuffer()
    XLSX.openxlsx(io, mode="w") do xf
        sheet = xf[1]
        XLSX.writetable!(sheet, collect(DataFrames.eachcol(df)), DataFrames.names(df))
    end
    return take!(io)
end

test_file = "test.csv"
gen_fake_csv(test_file, 100000)
df = CSV.read(test_file, DataFrame)
@time create_excel_file(df) # compile the method
@profilehtml create_excel_file(df) # profile

```

---

_[View the full topic](https://discourse.julialang.org/t/sending-excel-file-as-response-is-slow/116705)._