Batch macro to read 1000+ .xlsx files with varying schemas

I have a flow with a Directory input to read in all files from a network folder, I use a batch macro to find all sheet names across all of the files & then another batch macro to bring the data into the flow from each of those files, sheets.

Ideally I'd like to create a unioned view of the data BUT I'm struggling with the variances in format. All submissions should be the same format but various customers don't / won't comply. I'm happy to reject a number of submissions but some are simple fixes such as, the data doesn't start on row 1 so I need to look for row headings OR row headings are slightly different.

These are examples of some of the variances I'd like to fix. The batch macro which picks up all data from all sheets is failing when I bring it into my flow. The actual headers should be Distributor Account Number, Distributor Name, Date of Purchase, Customer Name, Material Reference Number & Quantity Sold but if this appears on rows 1, or even on rows 6,7 etc I want to start my data collation from there and if the headers say something like Dist Acc No, I want to accept that.

example file2.xlsx

example file3.xlsx

example file1.xlsx

Datasets

Batch Macro

Dynamic Processing

Accepted answers

caltang

Here's how I would do it @LizPerry

The concept is to use a batch macro to load the files in from a directory to get their sheet names. You can then filter for the sheets you want if they follow a pattern, for my case, I just used all your sheets.

The second step is to load the data from said sheets into the canvas using the same macro, but with a twist. You can see that I made the formula tools before the same macros to get the List of Sheet Names, and to replace said List of Sheet Names so that my macro can read the sheet info and the sheet data inside all in 1 go with a change of formula.

Then, we order the data and structure it, add a tile tool to identify each file called in, and then build a batch macro like so:

This batch macro basically standardizes the input, and orders it to become a proper table (removing null columns) and then positioning them to be dynamically renamed like what @EKasminsky said.

But here comes the extra twist... I added another control on top of the original control parameter, which takes in this text input information from the main workflow here:

Now, you can expand this list by adding to the text input yourself in the configuration panel on the left, or you can replace it with an .xlsx or .csv file as a mapping file. The idea here is to tell Alteryx that the headers within the datasets contain these key words, and as long as they contain these key words, we want them to be taken in as the 1st row before a dynamic rename.

This prevents scenarios like:

Where the first row is clearly not the header, but if we use the same method without accounting for the header's keywords, then we will get an issue.

Of course, I don't know how many keywords are there in your 1000 files, and I have controlled for them to get the first row that appears (in this case will be the main header).

Let me know if this works for you.

Best regards,

Calvin Tang

ForLizPerry.yxzp

All comments

EKasminsky

I'm not sure how helpful this is but I will put in my two cents.

I had a similar issue like this with multiple customers. What I did was I imported all of the customers directory of invoices separate from the other customers. So say I had Company A and Company B. Company A would be pulled in with one directory tool and then a macro to find sheet names, then one to import data. A separate process would be done with Company B. The reason I did it like this is because these individual companies tended to always send their invoices in the same manner. I could usually expect what the tab names and column names were.

I also use a Dynamic Rename for column names. This is just to remove any possible problem spaces in column names.

All of this makes a bigger workflow... but might help reduce the number of issues you are seeing or help.

caltang

Here's how I would do it @LizPerry

Then, we order the data and structure it, add a tile tool to identify each file called in, and then build a batch macro like so:

This batch macro basically standardizes the input, and orders it to become a proper table (removing null columns) and then positioning them to be dynamically renamed like what @EKasminsky said.

But here comes the extra twist... I added another control on top of the original control parameter, which takes in this text input information from the main workflow here:

This prevents scenarios like:

Where the first row is clearly not the header, but if we use the same method without accounting for the header's keywords, then we will get an issue.

Of course, I don't know how many keywords are there in your 1000 files, and I have controlled for them to get the first row that appears (in this case will be the main header).

Let me know if this works for you.

Best regards,

Calvin Tang

ForLizPerry.yxzp

caltang

Oh btw, the first directory tool you can make it like mine:

Store all your example files in a folder and load it from there.

flying008

Hi, @LizPerry

01- I assume you have already inputted all the data using batch macro [by field column position rather than by name].
02- Use keywords to identify the desired field name row and distinguish it from other data rows and rows to be deleted.
03- Filter out useless lines (such as the data rows above the actual field names).
04. Transpose by file name and identifier columns to facilitate unified field matching later.
05- Extract the first row based on the ID column, which is the row containing the actual field names.
06- Transpose.
07. Filter out empty lines, i.e., lines that are not common names.
08. Extract unique values, which serves as a preliminary step for manually handling irregular field names. (for below Step 09)
09. Manually match all ambiguous and partially matching field names; this step is crucial. ★★★

Value	Map
COGS	COGS
Cust Name	Customer Name
Customer Name	Customer Name
Customer name	Customer Name
Customer No	Customer No
Distributor Account Number	Distributor Account Number
Distributor Name	Distributor Name
Material Ref Number	Material Ref Number
Material Reference Number	Material Ref Number
Product	Material Ref Number
Online	Online
Date	Purchase Date
Date of Purchase	Purchase Date
Qty	Sold Qty
Quantity	Sold Qty
Quantity Sold	Sold Qty

10. Match the Map with the original field column names and join them into the transposed data from Step 4.
11 - Filter out empty rows, i.e., non-common name rows.
12 - Filter out the original first row, which is the header row containing column names, to prepare for the next Cross Tab step.
13- Group the data rows for identification, preparing for the next Cross Tab step.
14- Use the Cross Tab tool to restore the data to its original structure.

录制_2025_12_18_15_20_46_341.gif

Input
FileName	F1	F2	F3	F4	F5	F6	F7	F8
example file1.xlsx\|\|\|'Data$'	Customer No	Customer name	Date	Quantity	Product
example file1.xlsx\|\|\|'Data$'	123	Example Cust	01/07/2025	1	ABC123
example file1.xlsx\|\|\|'Data$'	123	Example Cust	07/07/2025	1	ABD456
example file1.xlsx\|\|\|'Data$'	123	Example Cust	31/07/2025	1	ABE789
example file1.xlsx\|\|\|'Data$'	123	Example Cust	29/07/2025	1	DF455
example file1.xlsx\|\|\|'Data$'	123	Example Cust	23/07/2025	1	DREW
example file1.xlsx\|\|\|'Data$'	123	Example Cust	23/07/2025	1	LKOO
example file1.xlsx\|\|\|'Data$'	123	Example Cust	30/07/2025	1	JHUS
example file1.xlsx\|\|\|'Report$'	Month:	25-Jul		testemail@test.com
example file1.xlsx\|\|\|'Report$'
example file1.xlsx\|\|\|'Report$'	Distributor Account Number	Distributor Name	Date of Purchase	Cust Name	Material Ref Number	Qty	Online
example file1.xlsx\|\|\|'Report$'	1234	Example Dist	01/07/2025	Example Cust	ABC123	1	No
example file1.xlsx\|\|\|'Report$'	1234	Example Dist	07/07/2025	Example Cust	ABD456	1	No
example file1.xlsx\|\|\|'Report$'	1234	Example Dist	31/07/2025	Example Cust	ABE789	1	No
example file1.xlsx\|\|\|'Report$'	1234	Example Dist	29/07/2025	Example Cust	DF455	1	No
example file1.xlsx\|\|\|'Report$'	1234	Example Dist	23/07/2025	Example Cust	DREW	1	No
example file1.xlsx\|\|\|'Report$'	1234	Example Dist	23/07/2025	Example Cust	LKOO	1	No
example file1.xlsx\|\|\|'Report$'	1234	Example Dist	30/07/2025	Example Cust	JHUS	1	No
example file2.xlsx\|\|\|'Page 1$'	Distributor Name	Distributor Account Number	Date of Purchase	Customer Name	Material Reference Number	Quantity Sold	COGS	Online
example file2.xlsx\|\|\|'Page 1$'	Example dist 2	12459	2025-11-3		twrweq	4	33.99	No
example file2.xlsx\|\|\|'Page 1$'	Example dist 3	12987	2025-11-3		potott	4	24.21	No
example file3.xlsx\|\|\|'Sheet1$'	Distributor Name	Distributor Account Number	Date of Purchase	Customer Name	Material Reference Number	Quantity Sold
example file3.xlsx\|\|\|'Sheet1$'	example dist in town A	777889	2025-12-1	A customer	7899988	1
example file3.xlsx\|\|\|'Sheet1$'	example dist in town B	777888	2025-12-2	A customer	655888	1


Output
FileName	Tile_SequenceNum	COGS	Customer_Name	Customer_No	Distributor_Account_Number	Distributor_Name	Material_Ref_Number	Online	Purchase_Date	Sold_Qty
example file1.xlsx\|\|\|'Data$'	1		Example Cust	123			ABC123		01/07/2025	1
example file1.xlsx\|\|\|'Data$'	2		Example Cust	123			ABD456		07/07/2025	1
example file1.xlsx\|\|\|'Data$'	3		Example Cust	123			ABE789		31/07/2025	1
example file1.xlsx\|\|\|'Data$'	4		Example Cust	123			DF455		29/07/2025	1
example file1.xlsx\|\|\|'Data$'	5		Example Cust	123			DREW		23/07/2025	1
example file1.xlsx\|\|\|'Data$'	6		Example Cust	123			LKOO		23/07/2025	1
example file1.xlsx\|\|\|'Data$'	7		Example Cust	123			JHUS		30/07/2025	1
example file1.xlsx\|\|\|'Report$'	1		Example Cust		1234	Example Dist	ABC123	No	01/07/2025	1
example file1.xlsx\|\|\|'Report$'	2		Example Cust		1234	Example Dist	ABD456	No	07/07/2025	1
example file1.xlsx\|\|\|'Report$'	3		Example Cust		1234	Example Dist	ABE789	No	31/07/2025	1
example file1.xlsx\|\|\|'Report$'	4		Example Cust		1234	Example Dist	DF455	No	29/07/2025	1
example file1.xlsx\|\|\|'Report$'	5		Example Cust		1234	Example Dist	DREW	No	23/07/2025	1
example file1.xlsx\|\|\|'Report$'	6		Example Cust		1234	Example Dist	LKOO	No	23/07/2025	1
example file1.xlsx\|\|\|'Report$'	7		Example Cust		1234	Example Dist	JHUS	No	30/07/2025	1
example file2.xlsx\|\|\|'Page 1$'	1	33.99			12459	Example dist 2	twrweq	No	2025-11-3	4
example file2.xlsx\|\|\|'Page 1$'	2	24.21			12987	Example dist 3	potott	No	2025-11-3	4
example file3.xlsx\|\|\|'Sheet1$'	1		A customer		777889	example dist in town A	7899988		2025-12-1	1
example file3.xlsx\|\|\|'Sheet1$'	2		A customer		777888	example dist in town B	655888		2025-12-2	1

*********

BTW, Due to security policy restrictions, I am unable to upload any. yxmd files, so please do it yourself,

LizPerry

Thanks @caltang , I'm working through your flow trying to understand and adapt it. Thanks for helping!!

Quick Links

This months top contributors

atcodedog05 19598

Qiu 15881

binu_acs 15708

MarqueeCrew 13708

apathetichell 13703