Hi, I just started with SQLITE.

I made a table with import
now I try to made a other small table with read a SQL file. I try with read but it didnt work.

if I try : sqlite> .read Source.sql
I have as result : Error: cannot open "Source.sql"

if I try : sqlite> .read "C:UsersGaelDocumentsProjet - SQLSource.sql"
I have as result : Usage: .read FILE

That mean, SQLite do not understand nothing.

Did you know how to do this ? I'm looking for in Internet and nothing is explain on this small aspect...

thanks for attentions

if I try : sqlite> .read Source.sql
I have as result : Error: cannot open "Source.sql"

This means that the file Source.sql was not found in the current directory. Since it appears that you are using Windows (based on your attempted use of DOS style pathnames) you can find out what the current directory contains by issuing the command: .system dir.

f I try : sqlite> .read "C:UsersGaelDocumentsProjet - SQLSource.sql"
I have as result : Usage: .read FILE

You seem to have forgotten that you need to escape your escape character if you use them as path separators. Furthermore, if your "filename" (or path segment) contains delimiters (eg, space), then you must put the filename in quotes.

If you wish to use the fullpath then you must specify the filename correctly. Any of the following is a correct DOS-style path:

.read "C:\\Users\\Gael\\DocumentsProjet - SQL\\Source.sql"
.read 'C:\\Users\\Gael\\DocumentsProjet - SQL\\Source.sql'
.read "C:/Users/Gael/DocumentsProjet - SQL/Source.sql"
.read 'C:/Users/Gael/DocumentsProjet - SQL/Source.sql"

It would probably be perspicacious, however, to simply execute the sqlite3 executable so that the current directory is correct, then you do not need to fiddle about with path specifications.

Are you sure you started 'sqlite3' in the same directory as the file 'Source.sql'?

Maybe try this:

.header off
.once test.sql
VALUES('SELECT 1');
.read test.sql

This will write out a file - which contains some SQL - then read it back in again. Assuming you can write to the file 'test.sql', this should work.

What's the point of the experiment? Once you have confirmed that you are able to write and then read 'test.sql' - make sure you file 'Source.sql' exists in the same directory.

  if I try : sqlite> .read "C:\Users\Gael\Documents\Projet - SQL\Source.sql"

I have as result : Usage: .read FILE

When I try the same, I get: Error: cannot open "C:UsersGaelDocumentsProjet - SQLSource.sql"

This is as I expect because the single, double-quoted argument to .read has backslash de-escaping applied before it is given to the .read command.

That mean, SQLite do not understand nothing.

Are double-negatives used commonly in your native language? To the extent that a program can be sensibly said to understand anything, it is true that SQLite understands more than nothing. Yet I doubt you intended to affirm this.

Try using single-quote characters around your .read argument. This will prevent the backslashes from being interpreted by the argument parser.¹ And take care that your 'Source.sql' file is in a subdirectory named 'Projet', or correct the spelling of that subdirectory name.

In the future, show what you submitted to the CLI, formatted per the markdown rules, so that others can see what is not working for you.

1. ^{^} See the section on Dot-Command Arguments here.