Spreadsheet Fields In Text: Fill & Download for Free

There are a few good reasons to use a well-defined serialization format.Text parsing and extraction is slow and burns a lot of CPU (especially at scale) compared to modern serialization schemes such as Avro (software), Google Protocol Buffers, etc. While modern languages and runtimes are smart (e.g. JIT optimization) regular expressions, especially those with backtracking, can bake down to complex state machines producing lots of (CPU) branching and cache churn. Poorly written code (which is stupidly common) can effect many temporary objects (or variables) that must be dealt with (e.g. GC'd).Working in text is actually a serialization language that is just poorly defined. When you have a "simple tab separated set of fields with newline separated records," you are really defining an ad hoc serialization format that is hard coded in a library or application. It's a "magic format" that only that code knows about. This leads to brittle code that is intolerant of surprisingly common operations such as adding or deprecating fields in the dataset. As parsing code is hard coded, it's impossible to discern metadata information and build generic tooling for data.The text representation of data can be significantly larger than that of binary. The canonical example of this is the integer. Consider the size of the number 12345 in memory. This value fits well within a 16 bit type (with a max unsigned value of 65536). That's 2 bytes. The string form of the value 12345, however, is 5 characters, each of which is 1 byte (mumble mumble UTF-8 mumble sizeof(char) ...) for a total of 5 bytes. That's a 60% waste. Reading or writing 5 bytes to disk takes more than 2 bytes. Compression can help here, but remember the first item above; you're burning a lot of CPU to parse, so you may already be CPU bound, in which case compression will only hurt. It's worth noting that MapReduce jobs over compressed text fields is usually non-splitable, and when it is, it's a giant pain in the ass (i.e. indexed LZO files).Representing complex types in text sucks. Yes, even in JSON.CSV is not (usually) CSV. I dare you to see how many of your "CSV files" are actually CSV (Common Format and MIME Type for Comma-Separated Values (CSV) Files). Note the introduction paragraph from that document (which was written in 2005):The comma separated values format (CSV) has been used for exchanging and converting data between various spreadsheet programs for quite some time. Surprisingly, while this format is very common, it has never been formally documented.Do yourself a huge favor and get into Avro (or whatever your favorite format is) in the first job you run on data. Better yet, write data from source applications and systems in a structured format from the beginning. It's not as scary or hard as you might expect and you will see benefits: well defined contracts and schemas, the ability to build general purpose tools, fewer parsing bugs, smaller disk footprints, reduced CPU consumption, and so on.With common formats like Avro, it's not usually necessary to write custom input formats or writables.

You might wanna give a try to Easy CSV Editor, too. (Yep, I'm its author!) . Some of the many features it has:* a native, spreadsheet like user interface, extremely easy to navigate and select* works with "broken" CSV files, like those which don't have the same number of fields in a row or fields which are not properly escaped* autodetects the format of the CSV, like which is the field separator* allows you to find & replace, in columns, rows or cells of your choice.* sort, merge, split - highly customizable* fill* "carve" into the last fields from a row so to be able to create CSV files which don't have the same number of fields in a row* undo, redo, move, cut, copy, paste and the rest of usual operation you would expect* text transformations* print, print to PDF, flexible PreferencesThe users' feedback is really important for me, and many of the features are developed because the users requested them - I consider this a crucial aspect of my application. In addition to US app’s reviews I encourage you to also check users' reviews from Germany and Netherlands, to see the very nice things they have to say about my app.

Apps script should be the simplest way for auto respond on form submit. You can do it in this way :go to spreadsheet with form responses - Tools > script editor and paste this codefunction myFunction(e) { MailApp.sendEmail({ to: e.values[1], subject: "Slack Invite", htmlBody: "Thanks for joining, also join us on slack <a href='slacklink'>here</a>" });}click on current project triggers button (clock button) and set on form submit trigger to myFunctionsave it, you get popup to grant permissions.Just change the subject and body as required and e.values[1] - here 1 indicates field number of your form, if email is 2nd field in your form then change it to 2 or appropriate field number.