<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: cheesecake</title>
    <description>The latest articles on DEV Community by cheesecake (@cheesecake_71738cfba6363c).</description>
    <link>https://dev.to/cheesecake_71738cfba6363c</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4257023%2Ff98e6571-947b-4124-815f-fc95da644262.png</url>
      <title>DEV Community: cheesecake</title>
      <link>https://dev.to/cheesecake_71738cfba6363c</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/cheesecake_71738cfba6363c"/>
    <language>en</language>
    <item>
      <title>Backfilling a year of credit card history into your budgeting app from PDF statements</title>
      <dc:creator>cheesecake</dc:creator>
      <pubDate>Sat, 21 Nov 2026 09:54:59 +0000</pubDate>
      <link>https://dev.to/cheesecake_71738cfba6363c/backfilling-a-year-of-credit-card-history-into-your-budgeting-app-from-pdf-statements-1onl</link>
      <guid>https://dev.to/cheesecake_71738cfba6363c/backfilling-a-year-of-credit-card-history-into-your-budgeting-app-from-pdf-statements-1onl</guid>
      <description>&lt;p&gt;Most budgeting apps (YNAB, Actual Budget, Tiller, Firefly III, Monarch, spreadsheets) can import a CSV, but card issuers usually only let you download the last few months as CSV. Anything older is a PDF statement. Here's a workflow for backfilling a year or more of credit card history without retyping it.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Get every statement PDF in one folder
&lt;/h2&gt;

&lt;p&gt;Download all the monthly statements and rename them consistently, e.g. &lt;code&gt;2025-01_visa.pdf&lt;/code&gt;. Sorting by name then gives you chronological order, which makes gaps obvious.&lt;/p&gt;

&lt;h2&gt;
  
  
  2. Convert each PDF to CSV
&lt;/h2&gt;

&lt;p&gt;For digital PDFs, a table extractor like &lt;code&gt;pdfplumber&lt;/code&gt; gets you most of the way:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;pdfplumber&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;csv&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;glob&lt;/span&gt;

&lt;span class="n"&gt;rows&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;
&lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;path&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="nb"&gt;sorted&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;glob&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;glob&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"*.pdf"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="k"&gt;with&lt;/span&gt; &lt;span class="n"&gt;pdfplumber&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nb"&gt;open&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;path&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;as&lt;/span&gt; &lt;span class="n"&gt;pdf&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
        &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;page&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="n"&gt;pdf&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;pages&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
            &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;table&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="n"&gt;page&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;extract_tables&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
                &lt;span class="n"&gt;rows&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;extend&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;table&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;

&lt;span class="k"&gt;with&lt;/span&gt; &lt;span class="nb"&gt;open&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"all_transactions.csv"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"w"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;newline&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;as&lt;/span&gt; &lt;span class="n"&gt;f&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="n"&gt;csv&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;writer&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;f&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;writerows&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;rows&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Card statements have their own quirks:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Dates without years.&lt;/strong&gt; "Dec 28" on a statement closing Jan 5 belongs to the previous year. Infer the year from the statement period, not from today.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Payments vs. purchases.&lt;/strong&gt; Some issuers show payments as negatives, others in a separate section. Decide on one sign convention (purchases negative is what most budget apps expect).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Foreign transactions&lt;/strong&gt; often span two lines (original currency + converted amount). Keep the converted amount.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Interest and fees&lt;/strong&gt; sometimes sit in a summary box rather than the transaction table. Don't miss them.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  3. Normalize to your app's import format
&lt;/h2&gt;

&lt;p&gt;Most apps want something like:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight csvs"&gt;&lt;code&gt;&lt;span class="nc"&gt;Date&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="nc"&gt;Payee&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="nc"&gt;Amount&lt;/span&gt;
&lt;span class="mi"&gt;2025&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="mi"&gt;01&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="mi"&gt;03&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="no"&gt;GROCERY&lt;/span&gt; &lt;span class="no"&gt;STORE&lt;/span&gt; &lt;span class="mi"&gt;123&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="mf"&gt;54.20&lt;/span&gt;
&lt;span class="mi"&gt;2025&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="mi"&gt;01&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="mi"&gt;05&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="no"&gt;PAYMENT&lt;/span&gt; &lt;span class="no"&gt;THANK&lt;/span&gt; &lt;span class="no"&gt;YOU&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="mf"&gt;500.00&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Use ISO dates (&lt;code&gt;YYYY-MM-DD&lt;/code&gt;) to avoid day/month confusion on import.&lt;/p&gt;

&lt;h2&gt;
  
  
  4. Reconcile before importing
&lt;/h2&gt;

&lt;p&gt;For each statement: previous balance + purchases + fees + interest - payments = new balance. If it doesn't add up, a row is missing or a sign is wrong. Fix that before it pollutes your budget.&lt;/p&gt;

&lt;h2&gt;
  
  
  5. Import oldest first and watch for duplicates
&lt;/h2&gt;

&lt;p&gt;Import month by month, oldest first. Where your app's own bank sync overlaps the backfill, delete the duplicates at the boundary.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; I'm part of the team behind &lt;a href="https://cardstatementcsv.com/" rel="noopener noreferrer"&gt;Card Statement CSV&lt;/a&gt;, which converts PDF credit card statements from any issuer (digital or scanned) into CSV. There's a free preview before you pay, no sign-up to start, files are deleted within 24 hours, and it's priced per conversion from $1. The manual route above works fine too if you have the time.&lt;/p&gt;

&lt;p&gt;How far back have you tried to backfill your budget? I'd like to hear what issuers made it hardest.&lt;/p&gt;

</description>
      <category>personalfinance</category>
      <category>python</category>
      <category>pdf</category>
      <category>productivity</category>
    </item>
    <item>
      <title>How to convert PDF bank statements to CSV (and import them into QuickBooks, Xero or hledger)</title>
      <dc:creator>cheesecake</dc:creator>
      <pubDate>Sat, 21 Nov 2026 09:50:45 +0000</pubDate>
      <link>https://dev.to/cheesecake_71738cfba6363c/how-to-convert-pdf-bank-statements-to-csv-and-import-them-into-quickbooks-xero-or-hledger-1kik</link>
      <guid>https://dev.to/cheesecake_71738cfba6363c/how-to-convert-pdf-bank-statements-to-csv-and-import-them-into-quickbooks-xero-or-hledger-1kik</guid>
      <description>&lt;p&gt;If your bank only gives you PDF statements (common for older months, closed accounts, or smaller banks), getting that data into a spreadsheet or accounting tool is surprisingly painful. Here are the options, roughly from cheapest to most automated, plus the gotchas that trip people up.&lt;/p&gt;

&lt;h2&gt;
  
  
  Option 1: Check for a hidden CSV/OFX export first
&lt;/h2&gt;

&lt;p&gt;Before converting anything, look in your online banking for "Download transactions", "Export", "OFX", "QFX" or "QBO". Many banks offer it but only for the last 90 days to 2 years. If it exists, use it: it's exact.&lt;/p&gt;

&lt;h2&gt;
  
  
  Option 2: Copy-paste or Excel's "Get Data from PDF"
&lt;/h2&gt;

&lt;p&gt;For a statement or two, Excel's &lt;strong&gt;Data → Get Data → From File → From PDF&lt;/strong&gt; can pull tables out of digital PDFs. It works well on clean layouts and badly on statements where descriptions wrap onto two lines or where each page repeats headers.&lt;/p&gt;

&lt;h2&gt;
  
  
  Option 3: Script it with a PDF table library
&lt;/h2&gt;

&lt;p&gt;If you're comfortable with Python, libraries like &lt;code&gt;pdfplumber&lt;/code&gt; or &lt;code&gt;camelot&lt;/code&gt; can extract tables from &lt;em&gt;digital&lt;/em&gt; PDFs:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;pdfplumber&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;csv&lt;/span&gt;

&lt;span class="k"&gt;with&lt;/span&gt; &lt;span class="n"&gt;pdfplumber&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nb"&gt;open&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"statement.pdf"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;as&lt;/span&gt; &lt;span class="n"&gt;pdf&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nb"&gt;open&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"out.csv"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"w"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;newline&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;as&lt;/span&gt; &lt;span class="n"&gt;f&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="n"&gt;w&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;csv&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;writer&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;f&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
    &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;page&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="n"&gt;pdf&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;pages&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
        &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;table&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="n"&gt;page&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;extract_tables&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
            &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;row&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="n"&gt;table&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
                &lt;span class="n"&gt;w&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;writerow&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;row&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Expect to spend time on cleanup: merged description lines, page headers in the middle of your data, debit/credit columns vs. signed amounts, and dates without a year.&lt;/p&gt;

&lt;p&gt;Scanned statements (images) need OCR first, which adds another layer of errors.&lt;/p&gt;

&lt;h2&gt;
  
  
  Option 4: A dedicated converter
&lt;/h2&gt;

&lt;p&gt;Tools built for bank statements handle the multi-line descriptions, running balances and per-bank layouts for you. Whatever you choose, check:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Privacy:&lt;/strong&gt; how long are files kept, and are they used for training?&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Accuracy:&lt;/strong&gt; does the closing balance reconcile? Always check opening balance + transactions = closing balance.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Preview before paying&lt;/strong&gt;, so you can see your own statement converted.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Importing the CSV
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;QuickBooks Online:&lt;/strong&gt; Banking → Upload transactions accepts CSV with 3 or 4 columns (Date, Description, Amount or Debit/Credit).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Xero:&lt;/strong&gt; Bank account → Import a statement, CSV with Date, Amount, Payee/Description.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;hledger / beancount:&lt;/strong&gt; write a CSV rules/importer once per bank and reuse it.&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; I'm part of the team behind &lt;a href="https://bankstatementscsv.com/" rel="noopener noreferrer"&gt;Bank Statements CSV&lt;/a&gt;, a converter for digital and scanned PDF bank statements. It shows a free preview before you pay, needs no sign-up to start, deletes files within 24 hours, and pricing is per conversion from $1. If your bank has a CSV export, though, use that first.&lt;/p&gt;

&lt;p&gt;Which banks have given you the worst PDF layouts? Curious what edge cases people run into.&lt;/p&gt;

</description>
      <category>python</category>
      <category>accounting</category>
      <category>pdf</category>
      <category>productivity</category>
    </item>
    <item>
      <title>A practical checklist for standardizing messy CSV files (with Python snippets)</title>
      <dc:creator>cheesecake</dc:creator>
      <pubDate>Sat, 21 Nov 2026 09:49:43 +0000</pubDate>
      <link>https://dev.to/cheesecake_71738cfba6363c/a-practical-checklist-for-standardizing-messy-csv-files-with-python-snippets-2599</link>
      <guid>https://dev.to/cheesecake_71738cfba6363c/a-practical-checklist-for-standardizing-messy-csv-files-with-python-snippets-2599</guid>
      <description>&lt;p&gt;Every team that imports CSVs from more than one source ends up with the same mess: the CRM export calls it &lt;code&gt;Phone&lt;/code&gt;, the spreadsheet from sales calls it &lt;code&gt;Mobile No.&lt;/code&gt;, dates arrive as &lt;code&gt;03/04/2026&lt;/code&gt;, &lt;code&gt;2026-04-03&lt;/code&gt; and &lt;code&gt;3 Apr 26&lt;/code&gt;, and somebody spends Friday afternoon fixing it by hand.&lt;/p&gt;

&lt;p&gt;Here's a practical checklist for standardizing CSV data before it hits your database or BI tool, with small Python snippets you can drop into a script.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Map column names to one schema first
&lt;/h2&gt;

&lt;p&gt;Decide on a target schema and map every incoming header to it. Normalize headers before matching so small differences don't break things:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;re&lt;/span&gt;

&lt;span class="k"&gt;def&lt;/span&gt; &lt;span class="nf"&gt;norm&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;h&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;-&amp;gt;&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;re&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;sub&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"[^a-z0-9]"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;h&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;lower&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;

&lt;span class="no"&gt;ALIASES&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="s2"&gt;"phone"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="s2"&gt;"phone"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"mobile"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"mobileno"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"phonenumber"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"contactnumber"&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="s2"&gt;"email"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="s2"&gt;"email"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"emailaddress"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"mail"&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="s2"&gt;"signup_date"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="s2"&gt;"signupdate"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"createdat"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"datejoined"&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;

&lt;span class="k"&gt;def&lt;/span&gt; &lt;span class="nf"&gt;map_headers&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;headers&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="n"&gt;out&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;
    &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;h&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="n"&gt;headers&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
        &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;target&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;names&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="no"&gt;ALIASES&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;items&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
            &lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="nf"&gt;norm&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;h&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="n"&gt;names&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
                &lt;span class="n"&gt;out&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="n"&gt;h&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;target&lt;/span&gt;
    &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;out&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  2. Parse dates explicitly, never by guessing per row
&lt;/h2&gt;

&lt;p&gt;&lt;code&gt;03/04/2026&lt;/code&gt; is March 4 in the US and April 3 almost everywhere else. Pick the format per &lt;em&gt;source&lt;/em&gt;, not per row, and fail loudly on anything that doesn't match:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;from&lt;/span&gt; &lt;span class="n"&gt;datetime&lt;/span&gt; &lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;datetime&lt;/span&gt;

&lt;span class="k"&gt;def&lt;/span&gt; &lt;span class="nf"&gt;parse_date&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;value&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;fmt&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;-&amp;gt;&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;datetime&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;strptime&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;value&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;strip&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;fmt&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;date&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;isoformat&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;  &lt;span class="c1"&gt;# always ISO 8601 out&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  3. Normalize phone numbers to E.164
&lt;/h2&gt;

&lt;p&gt;Strip formatting and store one canonical form. The &lt;code&gt;phonenumbers&lt;/code&gt; library handles country rules properly:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;phonenumbers&lt;/span&gt;

&lt;span class="k"&gt;def&lt;/span&gt; &lt;span class="nf"&gt;to_e164&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;raw&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;default_region&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;-&amp;gt;&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt; &lt;span class="o"&gt;|&lt;/span&gt; &lt;span class="kc"&gt;None&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="k"&gt;try&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
        &lt;span class="n"&gt;n&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;phonenumbers&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;parse&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;raw&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;default_region&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
        &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;phonenumbers&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;format_number&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;n&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;phonenumbers&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nc"&gt;PhoneNumberFormat&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="no"&gt;E164&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="n"&gt;phonenumbers&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;is_valid_number&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;n&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;else&lt;/span&gt; &lt;span class="kc"&gt;None&lt;/span&gt;
    &lt;span class="k"&gt;except&lt;/span&gt; &lt;span class="n"&gt;phonenumbers&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nc"&gt;NumberParseException&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
        &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="kc"&gt;None&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  4. Clean names and free text conservatively
&lt;/h2&gt;

&lt;p&gt;Trim whitespace, collapse double spaces and fix casing only when the whole value is upper- or lower-case. "McDonald" and "van der Berg" should survive.&lt;/p&gt;

&lt;h2&gt;
  
  
  5. Currencies: separate the number from the symbol
&lt;/h2&gt;

&lt;p&gt;Store amounts as decimals plus an ISO currency code (&lt;code&gt;1234.50&lt;/code&gt;, &lt;code&gt;EUR&lt;/code&gt;), not &lt;code&gt;€1.234,50&lt;/code&gt;. Watch for locales that swap &lt;code&gt;.&lt;/code&gt; and &lt;code&gt;,&lt;/code&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  6. Validate every row and keep the rejects
&lt;/h2&gt;

&lt;p&gt;Don't silently drop bad rows. Write them to a separate file with the reason, so whoever owns the source can fix it upstream.&lt;/p&gt;

&lt;h2&gt;
  
  
  7. Make it repeatable
&lt;/h2&gt;

&lt;p&gt;The real win is doing this the same way every time. Save the mapping and rules per source and run them on every new file rather than re-deciding each week.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; I'm part of the team behind &lt;a href="https://datastandardize.com/" rel="noopener noreferrer"&gt;Data Standardize&lt;/a&gt;, a web app that does the steps above without code: you define a template once (target columns plus formats), and every new CSV is mapped, normalized and validated against it. There's a free demo with no card required. If you'd rather keep it in code, the snippets above are a solid starting point.&lt;/p&gt;

&lt;p&gt;What's the messiest CSV format you've had to deal with? I'd love to hear the edge cases.&lt;/p&gt;

</description>
      <category>python</category>
      <category>datascience</category>
      <category>csv</category>
      <category>productivity</category>
    </item>
  </channel>
</rss>
