The Import screen
The header counts the imports this organisation has run. Below it are four sections.Have a workbook? Upload all of it at once
Choose an Excel workbook (.xlsx or .xls) with one tab per list — Sites, Assets, Parts — and it is
imported as one thing. The tabs are matched to the lists they hold, checked against each other, and
written in dependency order, so the assets tab’s “Site 2” finds the site the sites tab is about to
create. You see everything it will do before any of it happens, and it can be undone as one thing.
Name the tabs after the lists they hold. A tab named after nothing the product recognises is left out
and said so, rather than imported as a guess.
What do you want to import?
For a single list in a single file, click the list. The lists are grouped the way Administration groups its screens, and within each group they come in dependency order — the lists a register’s cells name come before the register, because a file of assets that says “Site 2” needs Site 2 to exist. Import them in that order. What can be imported (for an organisation with every module switched on):
You see only the lists you may import: importing a list needs the Import Studio permission and
the permission to create that kind of record by hand. If you hold neither combination, the screen
says nothing can be imported by you. Lists for modules that are switched off do not appear.
Clicking a list opens a new import for it straight away and takes you to its page.
Mappings you have kept
Appears once somebody has kept a mapping (see step 6, keeping the mapping, below).
Search kept mappings finds a mapping by its name, and the filter button narrows the list by any
column.
Kept mappings are also what file exchange feeds import through. A mapping is
offered for keeping only when somebody chose at least one column by hand or answered an unfamiliar
value; one the studio worked out by itself it will work out by itself again.
Every import this organisation has run
The last fifty imports, newest first. Click a row to open it.Where an import has got to
An import’s page shows where it has got to in three steps, across the top: Upload, Review and Import. The step you are on is highlighted, with a spinner while the product is working on it, and a step that is finished is ticked. The steps are worked out from the import itself, so a reload, a second tab or a file finishing in the background can never leave them saying something that is not true.
Once every step is ticked the import is done: it can be undone while its window lasts, and any
follow-ups — invitations, stated statuses — appear below.
On the list of imports, each import’s State reads as a word: Received (waiting for or reading
its file), Read, Mapped, Checked, Importing, Imported, Failed (the reason is at
the top of its page) or Undone.
One import, step by step
The import’s page shows only what applies to where it has got to. The header names the file (or “New …” before there is one), with Every import to go back, the three steps, and these figures once the file has been read:1. Hand over the rows
Under The file, choose one of three tabs:
A pasted or typed grid is saved as a CSV and goes exactly the way a dropped file goes: the same
reader, judged in full before anything is written. Paste something else clears the grid.
No file yet? Blank template downloads a workbook with the columns this list accepts, what
each one wants, and what already exists for the columns that name other records — the categories,
sites or vendors you can refer to.
While the file is read, Reading the file says so with a spinner. The page updates by itself, and
nothing is lost by leaving it and coming back. If something goes wrong:
When the file has been read, every column whose heading and values match one of the list’s fields
exactly is mapped for you, and the file is checked at once. A tidy file — one exported from the
product’s own Blank template, or a workbook whose tabs carry the product’s own column names — goes
straight to the review without a mapping step. A file with any column the studio is unsure of waits for
you at step 3.
A column that names another record is also recognised by the name of the thing it refers to, not only by
the field’s label. That keeps files saved from earlier releases mapping when a label changes: a parts
file headed Warehouse, from before stores were called stores everywhere, is still matched to Store.
A workbook started from Upload all of it at once whose file never arrived shows The workbook
with a file input to try again.
2. Use a kept mapping, if one fits
If a mapping you kept before fits this file, You have mapped a file like this before lists the candidates, best first, each with how many of the file’s columns it knows, how many it maps, how many answers it carries and when it was last used. Use this mapping maps every column it recognises and brings its answers with it — the records it already knows this file’s names for. It is offered only before you map anything yourself.3. Say what each column means
What each column means lists every column in your file, with a count of how many are mapped.
The Import into list has up to three groups:
Columns the studio could not place. When some columns could not be placed from their values,
a box says how many and offers Read the headings. That asks the AI assistant to read the headings
of only those columns, and lists a suggestion and its reasoning for each — for example that “BRAND”
means Manufacturer. A suggestion is not a mapping: nothing is applied until you choose the field
yourself. Reading headings counts against your organisation’s daily AI allowance, so it happens only
when you click.
Below the table:
Check the file saves the mapping and checks every row. It writes nothing. For a sheet of a
workbook the button is Save this sheet instead: sheets are checked together, in import order,
because one sheet names records another has not created yet.
Once a file has been checked, the column table is folded away. Adjust columns, beside the verdict,
opens it again if you want to change what a column fills.
4. Answer the new values
New values in this file lists each value the file names that matches nothing you have — one line per value, not per row, so a 50,000-row file with 47 unfamiliar manufacturers is 47 questions. Each line shows the value, the column, which kind of record it should name and the column it was matched on, how many rows use it with example row numbers, and more than one of yours is called this when the match is ambiguous. What should it be offers:
Records that decide who can see what — sites, locations, stores and parent assets — and records that
need facts a file does not carry, such as a part number or a unit of measure, cannot be created from a
file. Add those by hand and check the file again.
Each answer is saved as you choose it. Check the file again applies your answers; the counts do
not change until you do.
Files the sheet names
Once you map a column to one of your file or photograph fields, a Files this sheet names panel appears above the problems. A cell names its files —seal.jpg, or
several separated by a semicolon — and the files themselves are added here. Each is attached to the
record its row becomes, exactly as if it had been attached on the form.
Add files takes the files one by one or as a zip, such as the one an old system exported its
attachments as; a zip is opened in your browser and its files sent up one at a time, with Adding
… of … while they go. You can also drop them on the panel. Each file added is listed with its
state — Ready, Uploading, Refused by the scan or Did not arrive — and not named by
any row where no cell asks for it. A line above the list counts the files the sheet names that have
not been added yet, and names them. The bin beside a file removes it.
Every file goes through the same scan and size limit as one attached on a form. After each round of
files the sheet is checked again by itself, because a row that said “no file called seal.jpg” may now
have it. Once the import has been written, the panel is read-only: its files are on the records.
5. Read what will happen
When nothing in the file has a problem, nothing needs approval, nothing will be updated and nobody chose a column by hand, the verdict is one line — Ready: 12 assets will be created. — with the Import button beside it. Otherwise What will happen (on a sheet of a workbook, What the … sheet will do) says how many rows will be created, how many existing records will be updated, and how many have problems. Rows with problems must be fixed or left out before the import can run. It can also show:
The review asks the questions the import itself would: a serial number already in the register or on
two lines of the file, a name that clashes with one you already have, a rule a record must satisfy, a
plan whose checklist has nothing published, or a row that names a record another sheet’s rows will only
send for approval. Each is listed as a problem on its row, in words that say why, rather than discovered
when the import runs.
A few lists fill in what a file usually leaves out. A criticality level or a priority imported without
a rank takes the next place after your highest, in file order. Checklist steps without a number are
numbered in row order, and a checklist that has never been published is published when its steps
finish importing, so a maintenance plan file imported next can copy them; a checklist already in force
keeps imported steps as a draft for somebody to read.
What is wrong, and how much of it groups the problems: Reason, Column (or the row as a
whole), Rows affected and For example row numbers. It counts the whole file.
Rows to fix lists only the rows that need somebody, in a grid you can type into. Correct a cell and
the file is checked again by itself a moment later — there is no button to press, and the counts and
the Import button follow. Your correction is kept beside what the file said, not over it, so the
original row stays on record, and a line beneath counts the rows you have corrected.
A row you cannot fix here — a machine the file lists twice, say — can be left out of this import:
When every row with a problem has been corrected or left out, the grid disappears and the Import
button becomes available. The first 200 rows that need you are shown; the problem list counts the whole
file.
6. Keep the mapping
At the bottom of What will happen:7. Import
Import n rows writes every row the review counted. It is disabled while any row still has a problem — hover over it to read Fix or leave out the rows with problems first — and when there is nothing to apply. People without the commit permission see Somebody holding the commit permission has to say go instead. Rows left out are not written. A large file shows Importing with progress while it is written in chunks. When it finishes, Done says how many records were created or applied and when, and how long the import can still be undone. Rows held for approval are counted apart, as sent for approval. The rows then shows each row’s Row, State, Problem and What the file said, and The rows with problems downloads any rows that were not written, as a spreadsheet with your own columns in your own order and the reason in the first column. The State of a row says in words what became of it:After the import
Two follow-ups appear only where they apply.Undoing an import
Undo this import (in the header of a committed import, while the window lasts) removes what the import created and puts back what it changed, so long as nobody has touched it since. Records edited or referenced by other records since the import are kept. Approval requests the import raised that nobody has decided yet are withdrawn, so a later approval cannot create a record the undo removed. A request an approver had sent back to the person who raised it, for revision, is the exception: only that person may withdraw it, so it is left with them and the Undone section says how many are still waiting to be corrected — each can still be sent on and create its record. The Undone section then reports how many were removed, put back, kept because they had been edited, and kept because other records point at them — and how many new records the import created from your “create it” answers were retired too.Importing a workbook
A workbook import has one page per sheet, plus a This workbook section on every sheet’s page:
A sentence above the table sums it up: how many sheets and rows, how many sheets still need their
columns chosen, how many rows will be created and how many have problems, or what was undone. Where your
organisation approves new records of some of these kinds, a note says how many rows will be sent for
approval rather than created, and that rows on other sheets naming one of them are listed as problems
until it has been approved.
A tidy workbook arrives mapped and checked. Choose the columns of any sheet that asks for them on its
own page with Save this sheet, then use the workbook’s buttons, which act on every sheet at once
from any of them:
A sheet of a workbook has no import button of its own; its verdict says Imported with the rest of the
workbook.
Who can use this
See roles.
File exchange
The same imports, run unattended from files that arrive on an SFTP server.
Data export
Getting a copy of everything you can see, the other way round.