Get information from GridFS

[FidelCastillo]

hi, I'm tring to use the GridFS functions, and I'm pretty sure that I'm doing something wrong.

This is the pymongo code that I'm tring to replicate in mongolite:

client[ = MongoClient()
db = client['MotorCalculo']
fs = GridFS(db, collection='MotorCalculo')

DF = list()

for grid_out in fs.find({ 'metadata.CodKey':'123'}, no_cursor_timeout = True):
    info = json.loads(grid_out.read().decode())
    DF.append(info['NameCol'])`

and this is the mongolite code that I'm using

conFiles <- mongo('MotorCalculo.files', 'MotorCalculo')
gridFS <- gridfs("MotorCalculo")
gridFS$list(filter = '{"metadata.CodKey":"123"}')`

but I'm not able to read the gridFs out file

I have this

> gridFS$list(filter = '{"metadata.CodKey":"123"}')
[1] date
<0 rows> (or 0-length row.names)

thanks in advance

[jeroen]

That seems correct. Do you see anything if you do not pass the filter argument?

[jeroen]

What exactly does your metadata look like? This seems to work for me:

# some example data
fs <- mongolite::gridfs("test")
fs$write('iris', serialize(iris, NULL), metadata = '{"author": "jeroen"}')
fs$write('mtcars', serialize(mtcars, NULL), metadata = '{"author": "jimmy"}')
fs$write('cars', serialize(cars, NULL), metadata = '{"author": "jimmy"}')

# shows files with author jimmy
fs$find('{"metadata.author": "jimmy"}')

# drop example data
fs$drop()

Are you sure you are using the correct db? What do you get for fs$find() without filters?

[FidelCastillo]

Sorry I've got a few days off of my work. On Monday l'll be there. But I think that I can get the data from the gridFS

[FidelCastillo]

hi again, and sorry for delay... too busy in the last weeks.

this is what I have

library(mongolite) library(data.table) library(jsonlite)

conFiles <- mongo('MC.files', 'MC') conChunk <- mongo('MC.chunks', 'MC') fs <- gridfs("MC")

When I connect using the gridfd function I don't get any information

fs$find('{"metadata.CodMeta" : "123"}') [1] id name size date type

<0 rows> (or 0-length row.names)

But when I try with the mongo function I get the _id ,

data.table(conFiles$find('{"metadata.CodMeta":"123"}', fields = "{\"_id\":1}" )) _id 1: 5a8db149de310e302c1a59be 2: 5a8db149de310e302c1a59c4 3: 5a8db149de310e302c1a59c8 4: 5a8db0c4de310e302c19cda9 5: 5a8db149de310e302c1a59d4

---

5419: 5a8db0c4de310e302c19cd8b 5420: 5a8db0c4de310e302c19cd8f 5421: 5a8db0c4de310e302c19cd95 5422: 5a8db0c4de310e302c19cd9b 5423: 5a8db0c4de310e302c19cd9f

If I try the the empty find I don't have the data either

fs$find() [1] id name size date type

<0 rows> (or 0-length row.names)

Thank again, I'm sure that I'm pretty close

[jeroen]

Ah you are using a custom prefix? Try this: gridfs("MC", prefix = "MC")

[francisliyy]

Hi, why I can't use "fs <- mongolite::gridfs("test")", it says "Error: 'gridfs' is not an exported object from 'namespace:mongolite'

[jeroen]

You need the development version of the package:

devtools::install_github("jeroen/mongolite")

[francisliyy]

I got this error:

ERROR: configuration failed for package ‘mongolite’

• removing ‘/Library/Frameworks/R.framework/Versions/3.4/Resources/library/mongolite’ Installation failed: Command failed (1)

[jeroen]

You need to give more output if you want me to help you.

[jeroen]

The CRAN version now has gridfs support so you no longer need to build from source:

install.packages("mongolite")